CF driver Va - Vk voltages in SSS consensus?

[ElectronAvalanche]

Quick update:

I installed the bipolar board today and form a technical point it works perfect. I get. +/- 160V with the 60V bias tap of my PT.

IMG_4126.jpeg

I have not yet listened to the amp but will report back if there is any perceivable difference.

The other pic shosws the old board for bias as per the handdrawn Japanese schematic. At least now I have peace of mind in regards to the CF driver voltages. Maybe NOS or certain brands of current tubes work fine with the original 002 bias circuit, but a CF tube failure as I witnessed (maybe the tube was bad and would have failed anyhow) can not happen anymore due to excessive high voltages.

Happy Camper, thank you Martin!

Electron

IMG_4128.jpeg

[rootz]

In either case Vg-k doesn't get out of hand, so I don't think there is much danger of arcing there. Low voltages on the Japanese schematic are possibly just due to Japan's 100 VAC line voltage. The 378V at the plate node would become 454V with 120V input.

I heard that explanation before. I'm not familiar with US mains voltages and how they developed over the years. But I see now the AA270 405V plate voltage was specified for 120 VAC line voltage. I initially thought that would be on 110V.

Martin, aren't those cathodes in the preamp CF's at 215VDC or thereabouts? That is higher than the datasheet allows for. Now all my Marshall are also higher and seem to be doing fine, so this all might be a moot point. But they do violate the datasheet in that respect, right?

[rootz]

Is it me, or is that 5uF cap on the 820r resistor backwards? That would be part of the mixer section.

[rootz]

And while I'm at it: schematic... Some obvious liberties taken: FET on relais, OD ratio and presence control. The wiring on the can cap is a bit strange: OD and clean pre are not on the last two PSU nodes. I think this is how it should be, as ODR008x looks exactly the same.

I'm not entirely sure how the feedback in the reverb is done. Also bright cap on reverb return is difficult to see. Value of the trimmer after the reverb is not visible either. I took 250k as a working value giving roughly the same signal at the input of the OD as the normal output from V1b would do.

Suggestions welcome.

[bepone]

i know what is direct coupled driver for the output tubes.. it is existing maybe 100 years.

every RF amp has it because output tube for RF cannot work in negative bias ,usually is working in positive bias.. for that we need A2 driver directly coupled.. there is no distortion, normal stuff.
It is providing driving output tubes with low impedance, because grid when pushed to positive it is having very low impedance.. also when signal arrived there, output in A2 area is even more linear

[martin manning]

aren't those cathodes in the preamp CF's at 215VDC or thereabouts? That is higher than the datasheet allows for. Now all my Marshall are also higher and seem to be doing fine, so this all might be a moot point. But they do violate the datasheet in that respect, right?

With the Dumble driver supply the cathodes are at -225V on cold start-up. Typical Fender/Marshall DCCF has the cathode idling at +200V. Some US data sheets say +/- 200V is the max limit, some Russian tubes are only rated +/- 100V. Both JJ and PSVane are rated at 180V, with no polarity specified. With the bipolar supply cold start is at ~120V.

[Richard1001]

I have been thinking about the powersupply and standby switch in the SSS002 and came up with a theory that i tested.

Rootz stated that there would be no room for the extra 50uF caps on the supply board, based on the digital gutshot he made. But when i look at the gutshot of the SSS004 - which has the same size pre amp boards - there seems to be plenty of room for a bigger supply board and the two extra caps.

sss3.jpg

In the SSS004 however the additional 50uf caps before the standby switch are not present. But there is also no DC coupled CF. ( No CF at all)

a few years back when i measured a negative voltage on the power supply caps with the standby switch engaged, i placed a reverse diode over the powersupply to drain the current coming from the CF's negative supply to prevent a negative charge on the powersupply caps.

Looking back, another solution would have been to keep a positive charge on the powersupply caps by a resistor across the standby switch. I tried this yesterday and it works very well. I used a 56K 5W resistor. The 100uF caps keep a charge of +/- 90 volts with the amp on standby. The cathode' s of the CF go from -46 to -51 in standby. (was -70 Volt with the reverse diode)

With the 56K resistor the amp is silent. No sound is produced, even with all volumes on 10. Current flow is 6,25mA

[martin manning]

Seems like a pretty good fix. I'm still of the opinion that the bipolar supply is a better option, since the operating Va-k is much more reasonable, and it does not pull the driver cathodes down below -200V on cold start-up like the ~440V negative supply voltage will. I looked up the McIntosh MC60, and it has resistors at the power tubes from g1 to k like the SSS (MC30 does not), which will cut the negative voltage applied to the driver tube cathodes in half on start-up. I assume the intent there was to protect the h-k insulation on the driver tubes.

[dorrisant]

Just thought I'd post this here... the way I have been setting this up. Of course this is just a section of an example.

No preamp board space taken by the components, easy enough to add to just about any layout.

Feel free to critique. Hope it helps.

[martin manning]

That's a nice compact layout. I see you have added a bleeder (the 1M) on the positive side of the power supply, and you have grid stoppers on the CF's. Your bias adjust is not fail-safe, but that would be easy to fix by following the SSS 002 schematic implementation.

[dorrisant]

Your bias adjust is not fail-safe, but that would be easy to fix by following the SSS 002 schematic implementation.

Can you explain what should be done?

[martin manning]

Instead of taking the bias off the pot wiper you would use the pot to make a voltage divider with a variable ground leg. In the SSS 002, I like a 75k off the -170V, then a 24k3, then a 10k pot wired as a variable resistor to ground. Bias comes from the junction of the 75k and the 24k3. If the wiper contact fails, the variable leg goes to max resistance and bias goes to the most negative voltage.

[Guy77]

deleted

[dorrisant]

Instead of taking the bias off the pot wiper you would use the pot to make a voltage divider with a variable ground leg. In the SSS 002, I like a 75k off the -170V, then a 24k3, then a 10k pot wired as a variable resistor to ground. Bias comes from the junction of the 75k and the 24k3. If the wiper contact fails, the variable leg goes to max resistance and bias goes to the most negative voltage.

How is what I've posted different from the doc that shows how to splice this circuit into M Hartman's SSS?

Is there a jumper under the 10k trimpot? I'm confused. I understand the voltage divider, I'm just not seeing it in what I based my layout upon.

[martin manning]

That is the same as what you have, and the same as HAD did on ODS circuits. Personally I think it's pretty rare for a bias pot wiper to lose contact, but it's a trivial change to make sure that if it does happen the bias won't go floating.