CF driver Va - Vk voltages in SSS consensus?

[TroyBoy]

Update:

I installed a 22k (2W) resistor in series between B+ and the CF driver anodes and dropped the voltage to 380V (B+: 435V), this calculates to 2.5mA for the CF driver. I measured the bias and the output tubes lost 1mA.

I do not hear much of a difference between the high CF driver voltage and the lowered one and assume this will not make much of a difference or only at extremely high volumes.

BTW: the plates measure -47V when on standby, and this voltages slowly drops to -27V over time. Of course putting the amp in operate mode (standby off) results in the measured 380V on the plates.

So now I think I can play the amp with less anxiety that the CF driver tube may malfunction. Installing a 12AT7 would probably be even more cautious, albeit I am not really sure 12AT7 have a higher max Va-Vk rating.

Electron

Another thing to go back to is the RCA data sheets for a 12AX7 showing average plate ratings out past 400V.

A lot of these tubes will handle the voltage if they're set up ok. Jim Kelley took advantage of this pushing 6V6's beyond what most would do.

A VHT 2150 KT88 power amp would be another schematic i'd like to take a gander at, as I think you'd find some similarities in there as well...

A lot of interesting and valid inputs on this topic, so I'm definitely following..!!!

[ElectronAvalanche]

Update on this topic.

I use my 002 as a extravagant power amp currently to test my double SSS preamp and the Pentode preamp (feeding the output of the mentioned preamps into the return stage of my 002’s internal Dumbleator). Which means that there is a lot of putting th amp into standby and off standby.

So today my 002 stopped working all of a sudden when flipping standby off. No smoke fortunately. The B+ fuse (1A Fast) blew.

After much troubleshooting using the current limiter and measuring voltages, itbecame clear that the CF driver tube (7025 Chinese made) blew. The stage with the defect tube drew a lot of current, all voltages were down by almost half. When flipping the standby to off, the CF tube would make a funny noise like the plates inside would move.

A new tube brought the voltages in spec again. I have not yet tested the amp with audio signal, but I hope this was all.

Summary: I will implement the Manning bi-polar supply for the CF. I still wonder why Dumble would stress the CF so hard with his circuit. Also I reckon without a B+ fuse (as the original? I am not sure about this) the defect could have been more catastrophic.

I also still wonder how all the commercial 002 clones hold up. Do the manufacturers get a lot of calls because CF tubes failed. Was my tube an extraordinary bad specimen? Do users simply change the tube and call it a day? Maybe it simply boils down to how often the standby is switched stressing the tube with each switching?

Best Regards,

Electron

[rootz]

I'd definitely go with Martin's bipolar supply. It is a big improvement over the hand drawn schematic. It doesn't really matter much how others do it. The setup with the anodes of the CF connected to the screens of the power tubes just does not make much sense if you build an amp from the ground up. Take for example a look at SSS001: there is no visible connection between the anodes of the driver and the screens of the power tubes. Also the PI and driver board is almost the same as in AN's SSS and the HPD. Those amps have a separate node for the CF and a dedicated PSU part for the negative voltages. Is it such a stretch then to think 002's driver supply might have more in common with those amps? Anyway, much, much safer on the driver tube.

In 001 it even looks like the CF driver anodes and reverb driver are on the same node, but it's difficult to see in just one gut shot.

[bepone]

I also still wonder how all the commercial 002 clones hold up. Do the manufacturers get a lot of calls because CF tubes failed. Was my tube an extraordinary bad specimen? Do users simply change the tube and call it a day? Maybe it simply boils down to how often the standby is switched stressing the tube with each switching?

manufactrers often doesnt have a clue about designing something, specially in dumble's teritory.. 90% of all are just copy paste, without inspiration, only to take some money ..they doesnt even know how to design fuse type

[bepone]

for any output tube cathode follower directly coupled, +150 -100V will be enough blind guess, also mosfet can easily work there

[ElectronAvalanche]

I'd definitely go with Martin's bipolar supply. It is a big improvement over the hand drawn schematic. It doesn't really matter much how others do it. The setup with the anodes of the CF connected to the screens of the power tubes just does not make much sense if you build an amp from the ground up. Take for example a look at SSS001: there is no visible connection between the anodes of the driver and the screens of the power tubes. Also the PI and driver board is almost the same as in AN's SSS and the HPD. Those amps have a separate node for the CF and a dedicated PSU part for the negative voltages. Is it such a stretch then to think 002's driver supply might have more in common with those amps? Anyway, much, much safer on the driver tube.

In 001 it even looks like the CF driver anodes and reverb driver are on the same node, but it's difficult to see in just one gut shot.

Thank you Rootz for your comments. So in essence one could again argue that there is an error in the Japanese schematic in regards to the CF supply? Or in the design of 002? Also I take it from what you write that 001 might have a different CF supply (as reflected by Bludotones HPD and AN SSS (I thougth the AN is based on 002).

Thank you, Electron

[ElectronAvalanche]

I also still wonder how all the commercial 002 clones hold up. Do the manufacturers get a lot of calls because CF tubes failed. Was my tube an extraordinary bad specimen? Do users simply change the tube and call it a day? Maybe it simply boils down to how often the standby is switched stressing the tube with each switching?

manufactrers often doesnt have a clue about designing something, specially in dumble's teritory.. 90% of all are just copy paste, without inspiration, only to take some money ..they doesnt even know how to design fuse type

Hi Bepone,

Thank you for your comment. Indeed that could be the case. Copy and Paste. i must admit that this is also my approach since sadly I lack a lot of deep knowledge on tube electronics.But I am willing to learn and thank all the guys on this forum with more know-how than me for sharing and educating. I will try the Manning approach. Never felt good about the 002 CF supply and today I learned the hard way that the japanses schematic 002 CF supply might be flawed. Even tubes back when Dumble built the amp were not manufactured for the high plate to cathode voltages I think.

Best Regards,

Electron

[bepone]

Thank you for your comment. Indeed that could be the case. Copy and Paste. i must admit that this is also my approach since sadly I lack a lot of deep knowledge on tube electronics.But I am willing to learn and thank all the guys on this forum with more know-how than me for sharing and educating. I will try the Manning approach. Never felt good about the 002 CF supply and today I learned the hard way that the japanses schematic 002 CF supply might be flawed. Even tubes back when Dumble built the amp were not manufactured for the high plate to cathode voltages I think.

No problem we are here to share the knowledge and enjoy the tube circuits without stressing about some particular voltage, current, material, capacitor type, creativity is unlimited! Also tube circuit must be in first line RELIABLE.. so easy to change some voltage to be in the safety voltage ballpark...Every day i see bad designing pcb boards carry on high voltages, this is a new standard, people thinking that this is 9V pedal or so, then they even offer for selling this and trying to create a "brand"?!

Also many amp designers (i see many on this forum, "hiding" the production line?) selling too expensive their amps, but without any safety certificate? Yes people, reading this, it is forbidden, and if you can hold soldering iron in the hand, this doesnt mean that you can put your amp on the market! There are zillion safety regulations, and costly. Want to save? It is not possible here.
So, back to "manufacturers".. What does this mean? This average "manufacture" cannot pass any industry standard or satisfy some safety certificate, and you ask some design? i think proper engineering start to die in end of eight-ies, those units were simply built to last forever.. look inside the top of the line Tektronix scope from 1987 per example (2465B)..and then look in modern pcb amp nowadays

[martin manning]

So in essence one could again argue that there is an error in the Japanese schematic in regards to the CF supply? Or in the design of 002?

The original CF driver design has a couple of flaws. The high +/- voltage for sure, and as Richard1001 discovered, it puts a negative voltage on the filters when on standby. https://ampgarage.com/forum/viewtopic.p ... 16#p416116
The bipolar supply has neither of those problems.

[erwin_ve]

Design fault or not, modern 12ax7 tubes can't handle the CF voltages as in the Sss002 schematic.
As discussed in a other topic it might have come from this design:

McIntoshMC30.jpg

The Bipolar supply(Thanks Martin!) is the way to go IMO.

Erwin

[dorrisant]

I have built a few amps that are gigged and played quite regularly over the past 4-5 years. All of which were built or converted to Martin's bipolar supply. None of them have had any tube failures or any other failures. Loud, punchy and clear as a bell, they all get along lot of attention out in public. I'm sure the extra attention is due to what the cathode driver adds. The fact that they are bulletproof under hard use is no doubt due to Martin's supply circuit. Can't say enough good things about it, been suggesting it for years now.

[ElectronAvalanche]

Thnak you Martin and Erwin!

Indeed the Mc30 looks to be the template. Richard already pointed to that some while ago. Looking for MC30 and tube failure posts in the interwebs revealed that failure of the 12X7 CF in the Mc30 (with subsequent redplating of the power tubes) is not unheard of. Yet MC30s have been used for many many decades and are still revered. Richard pointed out (as does Merlin) that the max. Plate voltage (Va0) as per specs (old tube specs???) is 500V.

So maybe NOS or OS 7025 will last. Maybe current chinese 12AX7 might last in the CF position. Tony reported switching to 12AT7 due to his 002 eating AX7 too often. Maybe the MC30 without a Standbye-Switch does not stress the AX CF as much? I would assume that applying B+ whilst having -70V on the cathodes will get you close to Va0? B+ is 460V briefly in my amp and settles to 440V.

Richards Zener trick might be also something to consider, lowering the B+ to the CF by 120V.

@Erwin: did you implement the bi-polar cf supply in your amp? Who else out there, if I may ask?

Thanks,

Electron

[ElectronAvalanche]

I have built a few amps that are gigged and played quite regularly over the past 4-5 years. All of which were built or converted to Martin's bipolar supply. None of them have had any tube failures or any other failures. Loud, punchy and clear as a bell, they all get along lot of attention out in public. I'm sure the extra attention is due to what the cathode driver adds. The fact that they are bulletproof under hard use is no doubt due to Martin's supply circuit. Can't say enough good things about it, been suggesting it for years now.

Thanks for the feedback Dorrisant! That is quite the testament. I will order all the necessary parts this week.

Electron

[Richard1001]

The CF in the SSS 002 is in fact a copy of CF in the McIntosh MC-60 amplifier. The voltage chart of this McIntosh amp shows 430 volt on the plates of the 12ax7 and -45 volt on the grids and cathode. Some of these amps are still working today with the original CF tube. The McIntosh originally uses a Telefunken ECC83 (=12ax7) in that position.

When I was building my SSS, I also was worried to about the high voltage on this tube and tried to lower it in different ways. All of them made the amp sound weaker.
So I just ended up using the original schematic and high voltage.

And there really is no problem at all. No excessive wear of the tube, nothing blowing up. It just works fine.

The cheap Sovtek tube with box shaped anodes works in my amp. I tested the amp with those when i finished it. I changed tubes afterwards and just bought a NOS Telefunken for the CF. For the last two years the amp still works and sounds the same as the moment I put it in there. I do use an open shield on this tube. It gets hot.

[ElectronAvalanche]

The CF in the SSS 002 is in fact a copy of CF in the McIntosh MC-60 amplifier. The voltage chart of this McIntosh amp shows 430 volt on the plates of the 12ax7 and -45 volt on the grids and cathode. Some of these amps are still working today with the original CF tube. The McIntosh originally uses a Telefunken ECC83 (=12ax7) in that position.

When I was building my SSS, I also was worried to about the high voltage on this tube and tried to lower it in different ways. All of them made the amp sound weaker.
So I just ended up using the original schematic and high voltage.

And there really is no problem at all. No excessive wear of the tube, nothing blowing up. It just works fine.

The cheap Sovtek tube with box shaped anodes works in my amp. I tested the amp with those when i finished it. I changed tubes afterwards and just bought a NOS Telefunken for the CF. For the last two years the amp still works and sounds the same as the moment I put it in there. I do use an open shield on this tube. It gets hot.

Thank you Richard!

So my amp worked flawlessly since 2014. Granted it did not see as much use over all those years as within the last couple of weeks when, as said, I used it as a power amp for testing my Double SSS preamp build, Pentode preamp and ODS preamp build. When I would swap a part in one of the preamps, in the meantime I would switch 002 to standby and on again for testing the preamp and this would go on multiple times. So maybe that led to the failure of the CF? Or maybe it is just coincidence?

Martins bipolar would solve any potential problems, but it would of course leave the original schematic.

Electron