Overdrive Reverb/Steel String Singer Hybrid

[beasleybodyshop]

UPDATE: Revised schematic later in thread

Well, It's been a minute since I've been here - thought I would share a kitchen sink type project ive been working on for some time now.

I had this idea to combine a few different Dumble amps l really loved - Namely Overdrive Reverb #60 and the Steel String Singer. So I started to slowly piece together a schematic and layout.

Features:

1. Preamp topology/voicing of ODR#60
2. SSS style Step Filters, using Martin Manning's revised filter design.
3. 3 Tube Dumble Style Reverb circuit.
4. Post PI driver, using the 60v bias winding as a bipolar supply (also as per Martin Manning's design)
5. Independent Reverb controls for clean/OD channels
6. Replace FET circuit with Echoplex EP3 preamp.
7. PAB
8. Solid State effects loop
9. 100W output with pair of KT120s/6550s


Not really a lot of new ideas here, but I thought it would be a fun challenge to implement in one box.

I ended up using a 100W project transformer from ClassicTone before they closed out - running 560vDC Plate voltage. Using a Mosfet B+ dropper to bring the screens down to half potential.

Also using a 15v Tap on the PT to run the Relay supply, as well as the 9v input on my Echoplex preamp pcb. It has a charge pump on the board to bring it up to 21vDC.

The potential complications are many, but the main design considerations i have at this point is where to place the SSS step filters in the circuit. I am going to attempt to place it first right before the Reverb wet/dry mixer as it is in SSS#2/#5. I've attached a somewhat complete schematic to this post, will hopefully update as I make progress.

If anyone sees anything weird or obviously incorrect with the schematic please let me know.

Thanks!

[beasleybodyshop]

Removed

[Reeltarded]

Neat thing! So, I'm sort of doing this but a lil different.

Digital reverb late in circuit after SS loop.

Using the original topo filter w/Fasel inductor. Filter insterted right after V1. I think late circuit filtering setup might be a little too strongly represented in the output?

Can't remember whatelse.. oh! Using a 12au7 in place of FET with brights and cuts on a 6 position switch. Parallel, Split cathodes, tiny and big coupler with a balance control.. I think. Might break the pair out to one per channel. This is going to be a very low gain input driver.

Nice to 'see' you!

[bepone]

there will be too much dissipation on the power FET for the screens, and will blow very fast.

it is better to do find/order/put new trafo / with separate winding for the screens, lets say 200-240VAC, 0.3-0.5A, and then filter out desired voltage for g2, then will be almost no dissipation on the FET, but also FET no needed then.

[bepone]

paralleling like this 12AT7 which is made for osclillator use is never good..error which is copied everywhere almost 60 years.

need to be through separate cathode resistors and grids via separate grid stoppers, only anodes can be connected together....

[bepone]

feedback resistor 2k7 is maybe too low for the start, i would start with 5k6 connected to 4 ohms tap and then try to lower it depending of the quality in bass

[bepone]

using 12AX7 for the cathode follower is wrong, and is in opposite from the desired idea of cathode follower.. CF must provide current when output tube start to conduct in positive bias, means... maybe 10, 20, 30mA and then impedance of output tubes is very low..
means poor ECC83, 12AX7 has no any drive capabillity and will fail there

using low impedance double triode there is the way to go which can supply some current and work with lower impedances.... possible ECC82 - 12AU7, or what D used

[rootz]

using 12AX7 for the cathode follower is wrong, and is in opposite from the desired idea of cathode follower.. CF must provide current when output tube start to conduct in positive bias, means... maybe 10, 20, 30mA and then impedance of output tubes is very low..
means poor ECC83, 12AX7 has no any drive capabillity and will fail there

using low impedance double triode there is the way to go which can supply some current and work with lower impedances.... possible ECC82 - 12AU7, or what D used

Wrong depends on the context here. IIRC Dumble used the poor 12ax7 as a CF before the output tubes. But I recon a ECC99, 12au7 or something like that would suit better if class AB is your goal. If just avoiding blocking distortion is the goal, then a 12ax7 should suit. Better even would be a MOSFET. But that would make the amp the Solid State Singer.

[bepone]

i'm always looking from technical aspect..design must to be up to some standard, quality, not "random".
using 12AX7 for CF follower for KT88 in class AB2 is like using screwdriver to produce distortion on the speaker (holes in membrane)

[rootz]

there will be too much dissipation on the power FET for the screens, and will blow very fast.

it is better to do find/order/put new trafo / with separate winding for the screens, lets say 200-240VAC, 0.3-0.5A, and then filter out desired voltage for g2, then will be almost no dissipation on the FET, but also FET no needed then.

I did regulate the screens of a 6550 pair in one of my power amps. I used a Maida regulator for it and also regulated the bias supply. At idle the FET dissipated around 1W IIRC. I gave it a decent heat sink because at max power dissipation was more around 20W! But it worked back then and still does. The power tubes are driven by MOSFETS and the amp does around 80W clean. B+ around 500V IIRC, but it has been a while since I touched the amp. Also had a JTM50 clone with a MOSFET voltage dropper for the screens. I ran that amp flat out all the time. It kept working flawlessly.

But again agreed, a separate transformer is nicer. I did that in yet another pair of power amps where I used a 230-230V toroidal which are relatively cheap. Not cheaper than a simple regulator or MOSFET like in your design though.

[martin manning]

there will be too much dissipation on the power FET for the screens, and will blow very fast.

Seems to have worked fine here: https://ampgarage.com/forum/viewtopic.p ... 78#p440778

[bepone]

there will be too much dissipation on the power FET for the screens, and will blow very fast.

Seems to have worked fine here: https://ampgarage.com/forum/viewtopic.p ... 78#p440778

looks to me that amp is not bullet proof, you probably know the screen current curve, how it is raising on low anode voltages. so , calculate the power on the mosfet, is it helthy or not.

if the mosfet is only for running amp in the no-signal condition then i can agree.
but if the mosfet heat dissipation doesn't allow to run amp full signal, so can be understood like limiting the output power, and you want this because you are installing higher wattage tubes in the amp ?

so why not instead, to run separate isolated screeen winding, there is no heat, there is no problematic FET (reliability + oscillation) in the circuit, there is nothing! and forget about the worries

[bepone]

I gave it a decent heat sink because at max power dissipation was more around 20W! But it worked back then and still does.

i have calculated the same 15-20W.. it is a lot of heat! needs 0.5-1 kg of aluminium, with heavy heat sink can work, yes, agree, ok

[martin manning]

looks to me that amp is not bullet proof, you probably know the screen current curve, how it is raising on low anode voltages. so , calculate the power on the mosfet, is it helthy or not.

The MOSFET has a current limiter, which drops screen voltage and current.

[rootz]

I calculated roughly 20W of dissipation IF the screens pull 100mA in my amp. But it does not pull that much current anytime. IIRC is was around 50mA RMS max. That is under 10V positive grid circumstances, so a good bit in class AB2. Give the fact that this amp has a puny ECC83 as a driver, I don't think it will come close to 20W dissipation. And then there is the current limiter.