Harp Switch in AB763 circuit

[TheTinMan]

I've built a modified BF AB763 Super Reverb with bias vary tremolo and both channels going through reverb and all of the gain stages. The Vibrato channel is virtually stock and sounds great to my ears.

I'm still working on the Normal channel though. The Bright switch on the Normal channel is intended to be a guitar/harp switch. I really like the tone on the guitar setting - it's intended to have a Brown Vibroverb feel to it. The B+ is a lot lower for the Normal channel than it is for the stock Vibrato channel (about 255 VDC IRRC).

I have a question about the harp setting though. Most harp amp circuits do not seem to have a cathode bypass cap on the first triode stage. My understanding is that this is to minimize feedback, among other things.

Looking at the attached preamp schematic, right now I've got one pole of the Bright switch swapping a 4.7uf (guitar) bypass cap for a 47uf bypass cap (harp). I plan to wire it up so that I can use a jumper to swap in the 47uf bypass cap, but I don't know how much gain is too much. Note that the other pole of the guitar/harp (Bright) switch already reduces the gain a bit in the harp setting.

The harp player who is going to use this amp has a Super Reverb and apparently uses the Vibrato channel at least some of the time. That's a good thing since my Normal channel has all of the gain stages a typical Vibrato channel has. He uses a Shure Green Bullet mic. That mic has a volume control, so maybe that's how he tames the gain???

If the father/son guitarist/harp player were local, I'd just wire it up as shown and get the harp playing son to test it out. Unfortunately, I live in Virginia, Dad is in Maine, and son lives in Massachusetts. I could wire it up without the bypass cap but all set up for it and modify the amp a few months down the road. However, I'd sure like some insight from any of you guys who've made harp players happy.

Thanks in advance for any insight,

Chip

[cbarrow7625]

I don't have an answer to your question but I did want to ask if there was a mistake in your schematic. On the cathode of V1B you have the top of the 82k resistor tied to the bottom (ground) of the 1k resistor. Looks like the jumper from the 82k to the 1k is unnecessary (and shorts the 82k resistor out of the circuit entirely). The way it is drawn Bright-2 wouldn't do anything.

[TheTinMan]

I don't have an answer to your question but I did want to ask if there was a mistake in your schematic. On the cathode of V1B you have the top of the 82k resistor tied to the bottom (ground) of the 1k resistor. Looks like the jumper from the 82k to the 1k is unnecessary (and shorts the 82k resistor out of the circuit entirely). The way it is drawn Bright-2 wouldn't do anything.

I must confess that my initial response was "Of course there isn't a mistake - it works in the amp!" Then I looked at the schematic again... OOPS! I have corrected the mistake in the drawing.

With Bright-2 open, there is a 47uf cap bypassing only 470 ohms of the combined 1.47k cathode resistor. With it closed, the 47uf and 10uf caps in series result in an 8.25uf cap bypassing the entire cathode resistor.

Thanks for catching the error.

Chip

[MWJB]

"I have a question about the harp setting though. Most harp amp circuits do not seem to have a cathode bypass cap on the first triode stage. My understanding is that this is to minimize feedback, among other things." Maybe when discussing smaller SE & push-pull amps but typically, big, fixed bias 6L6 amps DO have 1st stage cathode bypass caps, tolex Fender based circuits usually have a 2nd stage bypass cap too. These caps might be larger that 25uf for harp, maybe 100-220uf. In a big amp like this, a lack of 1st stage bypass cap is likely to dull off the amp/make it too loose sounding.

It's not often that you'll see someone using more than 1500 per triode for the input stage cathode, 1.9K might be the highest I'd go. I'd be looking for 160-180vdc on the input stage plate. Also try 820 for the second stage cathode, for a hotter tone?

I prefer 0.022 for the middle cap value.

You have reverb on both channels, I'd rather have the original dry normal channel. In any caseyou apper to be missing 2x220K mix resistors after the preamp coupling caps, before the reverb is tapped?

"Green Bullet" covers a multitude of sins, it doesn't relate to any single mic. Vintage CM/CR green bullets will want to see 5meg input loads, the later 520DX (dynamic) with volume control will want 1Meg max.

Try a 12AX7 in the PI slot, maybe even a 12DW7, and a larger than normal PI coupling cap (0.1uf?).

Try bypassing the NFB load resistor with 1uf cap for a presence boost, but really I would recommend a 6G12A style PI with presence pot.

For a 4x10" I'd recommend lower than normal power tube plate currents, say 7-15mA per tube, I usually recommend Sovtek 5881WXT in these applications. Try a solidstate recto too, with the lowered preamp voltage the amp won't sound too hard/sterile, but it will keep up with a band easier.

[TheTinMan]

MWJB - thank you very much for your input! There's more to this amp build than optimizing it for harmonica but I really appreciate all of your thoughts.

Making the second stage cathode resistor smaller is an interesting idea that I will try.

BTW the plate voltages in the Normal channel are in the 165-180 range. I've read many of your other posts

I'll also experiment with a cooler bias on the power tubes. Right now they're pretty hot (about 52ma plate & screen current combined). The test tubes are Chinese 6L6-GCs, but the amp will be delivered with Winged-C Svetlanas.

The Mid cap value was a compromise - with the Bright switch on this is supposed to sound like a Brown Fender for guitar.

The owner-to-be wants reverb on both channels. I actually had mixing resistors between the coupling caps of V1-B and V2-B and the 10pf/3.3meg reverb takeoff point. They reduced the signal significantly - I expected a bit of a voltage divider with the 1 meg grid return resistor for the reverb driver but the the 220k going to ground after the Reverb pot also appears to come into play there. Bottom line - I jumpered the mixing resistors and volume went up to what I expected. I posted on the Hoffman forum and was told to go back and measure the drop:

1.2 KHz signal:
53ma w/o jumper
150ma with jumper

Showed up very clearly on the oscilloscope too.

If tying the plates together is good enough for Kevin O'Connor, it seems to me that duplicate coupling caps let me have different B+ and plate voltages and isn't any worse.

The 520DX Green Bullet is the mic in question. It needs to see a minimum input load of 100K - that's why the input jacks for the Normal channel are not wired stock Fender. The Lo jack provides about 100k input load and almost a 50/50 voltage divider.

Thanks again for your thoughts!

Chip