Dumbleator Layout

[Vertigo]

Great looking layout now I just need to make some boards.

Mark

If you want the DIY file, let me know. I haven't had a chance to implement Martins corrections yet. My wife went back to work yesterday and I'm trying to manage my business with two kids at home. Another day or two and I'll put the correction up, along with trace masks for the boards

[jam-mill]

Some recent discussion on the send buffer and cable loss (https://ampgarage.com/forum/viewtopic.php?f=4&t=32162) got me looking into the frequency response. It's true that the buffer is not very well equipped to handle long cables with its 250k send pot, and that will be a problem if the buffer's output is run out to a pedal board. Attached below is a tweaked schematic that addresses that, and includes an alternate power supply using a current production Hammond transformer that will fit in a 1u rack box.

The lower value send pot is accommodated by raising the cathode resistor and tweaking the bias point. If that is all you are after, you can just change the pot, Rk, and Rbias, and leave everything else alone.

Dumbleator_Update_MPM.png

Martin,

The Hammond 262A12 transformer, according to the data sheet, does not appear to have a center-tapped filament supply. Therefore, won't grounding the tube socket at pin 9 cause problems?

-jack

[martin manning]

The Hammond 262A12 transformer, according to the data sheet, does not appear to have a center-tapped filament supply. Therefore, won't grounding the tube socket at pin 9 cause problems?

No, the two 6.3VAC filaments form a faux center tap, similar to using 2x 100Ω resistors to create a ground reference. As shown in the schematic, the two filaments are wired in series, with pin 9 being the common connection.

[Charlie Wilson]

I know this is way, way late to the party but I'm curious about how the wattage ratings of the two dropping resistors in the original layout were chosen. The 15k in the photo looks to be a 1.5w Q-Line and the 100k looks to be a .7w Q-Line.
CW

[martin manning]

I know this is way, way late to the party but I'm curious about how the wattage ratings of the two dropping resistors in the original layout were chosen. The 15k in the photo looks to be a 1.5w Q-Line and the 100k looks to be a .7w Q-Line.

I don't know how the resistor power ratings in the original were chosen, but here's what I'd do:

These are Class A stages, so we can consider the average plate current to be constant at the zero signal level. Let's also say that the reservoir caps will charge up right away, to 360V.

The 100k carries around 1.1mA and drops 100V, which is 0.12W. With the usual 2x factor a 1/4W would be big enough, but 1/2W doesn't seem extravagant given there will be some inrush current at the second filter.

The 15k will be handling the current for both stages, say 2.2mA, which is only about 0.07W. However, when you turn the power on, the 360V B+ will be across the 15k (the filter looks like a short), so it will momentarily be absorbing/dissipating something approaching 10W.

When fully charged, the first filter will have CV^2/2 [J] stored in it, and that will take five time constants 5RC [s]. The second filter will be storing charge too, but its RC is almost 7x the first filter, so I'll just ignore that. The average power [J/s] = [W] the 15k will have to absorb/dissipate over that interval will be V^2/(10R) = 360^2/150k = 0.864W. Doubling that and rounding up, I would use 2W.

This seems reasonable for the case shown, but not for cases where the first dropping resistor is smaller, like the circuit I posted above using the Hammond transformer, or my own compact D'lator, which has a 3k9. One difference is that the reservoir does not come up to the B+ value immediately, so the charging interval is longer, and the second filter draws more charge, but it still looks like lower value first dropping resistors should be higher wattage. Either that, or resistors can take significant abuse transiently.

[Charlie Wilson]

Thanks Martin. I suspect that maybe that is what Brandon chose to use, so that is what went on the layout. What about the wattage/voltage rating of the 10m LNFB resistors? I have some 1/4w that I like to use in amplifiers but I don't know if that would be a good idea in this application.
CW

[martin manning]

What about the wattage/voltage rating of the 10m LNFB resistors? I have some 1/4w that I like to use in amplifiers but I don't know if that would be a good idea in this application.

I don't see a problem with the power rating. There is no DC on those resistors due to the cap, and since there are two in series, the AC voltage on each is halved.

I think the analysis above on the dropping string wattage requirements is on the right track. It comes down to using a resistor with enough mass to pass the initial surge of current without too much temperature rise, which is going to be much more than the steady state dissipation requirement. It'd be interesting to look at that with a thermal camera.

[Charlie Wilson]

I guess ideal would be a 2w carbon comp? I believe one of the 10m LNFB does have DC on it. The one coming off the plate.
CW

[martin manning]

I guess ideal would be a 2w carbon comp?

Flameproof metal oxide is the preferred choice for power resistors. Maybe someone will say cc sound better ;^)

I believe one of the 10m wLNFB does have DC on it. The one coming off the plate.

It’s true that the 220n cap will be charged through that resistor, but after that the cap blocks it.

[Charlie Wilson]

No not sound. My understanding is that carbon comp handle voltage spikes(pulses) better than most.
CW

[jazzbass]

Hi all,

I need a suggestion for the Hammond transformer, which transformer should I look for for the European wall voltage of 220V?

Thank you

Franco

[martin manning]

I need a suggestion for the Hammond transformer, which transformer should I look for for the European wall voltage of 220V?

I don't know of any single transformer that has the specs you are after.

Another option would be to use two flat-pack transformers: https://www.hammfg.com/electronics/tran ... er/229.pdf These are quite small in size, and inexpensive.
I would use 229E12 for the heaters, and 229B230 for the plates. The 229B230 has 230VAC secondary voltage with 230VAC input at full load (50mA). The regulation is not very good on small transformers, so the voltage will be quite a bit higher at partial load.

I have a very similar Triad FP230-50 unit here, and I measure 258VAC on the secondary (series connected) at 110VAC input (primaries connected in parallel) with no load. That would result in 363 VDC using a full wave bridge, again with no load. You would of course connect the primaries in series for your 220VAC line voltage. You won't be using anywhere near 50mA, so the voltage will not sag much. If I add a 50k load resistor to the secondary, the voltage only drops to 254, and that would be 5.1mA, more than enough for the two triodes in the D'lator, and that produce just about 360VDC.

[erwin_ve]

Hi all,

I need a suggestion for the Hammond transformer, which transformer should I look for for the European wall voltage of 220V?

Thank you

Franco

Hi Franco,

You could use a Amplimo 1N242 for this. Here's a link: https://www.ringkerntrafo.nl/shop/high- ... -30va.html.
I used it in 4 dumblelators. The Ps rectifier and filter are a different, with a FWB rectifier: https://ampgarage.com/forum/viewtopic.p ... 92#p234192

Erwin

[martin manning]

You could use a Amplimo 1N242 for this...

Looks like a good option. What voltage did you end up with, and what is the line voltage level where you are?

[erwin_ve]

You could use a Amplimo 1N242 for this...

Looks like a good option. What voltage did you end up with, and what is the line voltage level where you are?

Line voltage was around 230v.
My voltages on B+ dropping string are: To recover triode=345v; to cathode follower triode= 268v.

This is the exact schematic for the Ps filter, the first two lytics were lowered to 47uF in a later stage.