Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: Joel on November 01, 2016, 09:38:57 pm
-
Hi all,
I just finished my build based on Mr Hoffman's single channel Deluxe Reverb. It sounds very good. But I have two issues.
1. The B+ is very high. 470VDC at node A. I'm using the Hammond 290BEX, and my wall voltage here in South Oz is 245Vac. JJ GZ34 rectifier tube. It's all pretty close to the data sheet at no load considering my high input voltage. It appears my rectifier is very, very efficient. Too efficient. I was expecting something in the 420VDC region. All my power supply caps are 500V. Is this high B+ going to be an issue? Simple way to lower the B+? I've got a bunch of 5Y3s lying around...
2. There is a weird noise coming from the reverb. It's like the sound they use in the movies of water dripping in a cave. It's not regular, but it is often. I've swapped my 12AT7 tubes. I'm wondering if it isn't the 470V on the reverb driver tube plates causing some sort of intermittent shot noise or something? (The schematic has reverb driver plate voltage coming from node B which is close to 470VDC). I believe most 12A_7s like plate voltages much less than this. I'm thinking of either moving the reverb driver to another B+ node or lowering the whole B+ as above.
-
I think you have precisely the right idea. Throw that 5Y3 in there. With any luck that should knock down your B+ 45 or so volts and you could be done, no muss, no fuss.
-
Are the output tubes biased cold? Biasing them hotter should load that B+ down a bit more. Not much you can do if you are already running hot though.
The reverb issue may disappear if the voltage decreases. Consider adding a dedicated B+ node for the driver rather than moving it to an existing node.
-
Thanks for the quick repies guys.
My power tubes are old secondhand 6V6's. Ones GE and the other Philips. Bias is 17.5mA and 14.5 mA, so pretty close to 70% of max dissipation for the GE at 470V plate voltage. I'm assuming 12W max power from these old used tubes.
I might try a 5Y3 first. Otherwise, I have two 22uF caps in parallel after the rectifier and before the choke. I may split this into two nodes, with a single 22uF into the choke and then add another RC stage before the two existing 22uF/10K RC filter stages (if this makes sense). Perhaps a 22uF/1K combo?
-
Using an NOS 5Y3 will definitely drop the B+. But a 5Y3 will be running near it's max current rating in this amp. Running as cool as you are should not be a problem though.
Maximum plate dissipation for a 6V6 is 14 watts. Your tubes are running at 59% and 49%. Kinda cool. Just try running them around 25mA long enough to see how much the plate voltage drops.
My PT is 330-0-330 @ 150mA(slightly beefier than yours), using GZ34, with bias set for 24mA, and plate voltage at 423V. Node A B+ is 428V.
If you do decide to create a separate node for the reverb driver I would suggest leaving the parallel 22s on node A and use an additional 22µF connected as you suggested. Don't hesitate to use a larger resistor to drop the B+ to the point you prefer.
You'll probably need to leave the tremolo circuit connected to node B.
-
Thanks Sluckey,
I appreciate your advice. Upping the bias to 22mA dropped the plate voltage to 455V (and the rest comparatively less along with it). I swapped the reverb driver to node D as a temporary test, and that worked to remove the noise. 450V was definitely too hot.
I think I'll try this next:
Nodes:
A B C D E
In _______470R 5W____ 4H____10k____10k____ Out
| | | | | |
22u22u 22u 22u 22u 22u
Adding the 470ohm/22u RC combo straight after the parallel input caps (or maybe 1K). I think this is what you were getting at Sluckey? This should drop the voltage from nodes B onwards by 40 or 50 Volts yes? I might have room to shoe-horn this in.
-
This is what I had in mind. This allows you to fine tune the Reverb Driver node without affecting nodes C and D.
-
Thanks Sluckey. That worked very nicely.
I've now got a question about the Tremolo. There is a quiet but distinct thump. It is independent of volume. After taking heaps of readings in my amp I've since read that this is normal for tube bias tremolo. BUT, I also measured a max of 320V on the Cathode of the Tremolo tube (V4B)! That's well above the heater to cathode rating. This can't be good for the tube. Could a pair of LND150's be a good drop in replacement?
-
Tubenit has a circuit similar to what you suggest. I don't know if he has actually tested the circuit but it looks good on paper. See attachment...
I only have one amp with your trem circuit My B+ is 388V and the voltage on the cathode of the CF is 204. I don't have any concerns about those voltages. I think I'd try moving the tremolo B+ source to the newly created node for the reverb driver before I added the MOSFET.
Lowering the B+ to the trem circuit may take care of the trem thump. Adjusting the bias will also take care of the thump. There is always a compromise between bias and trem operation in this type circuit.
-
Thank you, thank you, thank you Sluckey for your patience and advice! :worthy1:
Thank you Mr Hoffman for the single channel design.
This is a *beautiful* sounding amp. Guitar --> Amp... Magic.
My thump is either gone, or with the intensity turned up, just too low to hear until you get your ear right up against the grill cloth.
1. I ended up using a 22k/22u RC filter to the new B1 node to power the Reverb and Tremolo. It now sits at 368V.
2. Lowered the power tube bias a fraction. The new B+ is 460V and bias is set for 21mA (19mA on the lower one). These 2 things lowered the thump effect, but didn't eliminate it.
3. Installed a 1.73V clear red LED to the cathode of V4A did the trick. One day I might bring the LED to the front panel for the cool visual, but for now I'm just enjoying the amp.
I also changed one 0.01u cap to 0.022u to make the tremolo effect a bit slower.
Thanks again.
-
Nice looking amp and the bottle brush makes a good background. Do you have any gut shot pics?
-
I just realized I didn't take any post mods shots... buggrit. But here is one I took before I added the B1 node and Tremolo cathode LED.
-
Looks good on the inside too. Very neat work.