First fire up of the power supply. Brought it up slowly on the variac with no problems or magic smoke escaping. 500VDC on the plate node and the MOSFET regulator is working spot on to deliver 400VDC to the screens and PI feed. Something isn't quite right with the bias supply but I'm outta time for tonight. Getting there
EDIT: bias sorted - it was just a cold solder joint. I'm going to need to bring the range down on it though as it's waaay overkill.
Loop wrote:I’m currently shopping for a 1996 Red Dodge Viper with yellow wheels. Who gives a shit about taste?!
Adjusted the bias supply to a sensible range. We know that to get the tube to pretty much cut off we need a bias of:
-V_screen/µ_tube or -V_plate/10
KT88s have the lowest µ_tube of anything I'm likely to stick in at a value of 7.5. 6550s are slightly lower at 6.8 which isn't a great deal different. This means they'll need the largest range to bias up. In this case
V_bias = -400/7.5 = -53VDC or -500/10 = 50VDC
with the stock circuit I had something utterly ludicrous like -125VDC! To cull this I've added a resistor between the -ve output of the rectifier and the bias reservoir cap. In this case I used 47k, which forms a potential divider with the bias pot and the range protection diode to bring the raw bias down to a more sensible -60ish VDC. I need to tidy the wiring up after all this fudging around but the power supply stuff is done now.
Next up I need to wire up the power tubes and drill holes for the preamp tubes. After that I can get this thing making some noise.
Loop wrote:I’m currently shopping for a 1996 Red Dodge Viper with yellow wheels. Who gives a shit about taste?!
In the name of making little to no progress I'm considering binning this chassis off and building inside a 4U rack chassis instead. That removes the headache of building a box for it to live in. Good god I love procrastinating haha!
EDIT: nah, I'll just build this as a stand alone poweramp in it's current chassis to test a few preamps out and then I'll think about converting it to a full 4u rack amp when I know what works.
Loop wrote:I’m currently shopping for a 1996 Red Dodge Viper with yellow wheels. Who gives a shit about taste?!
Made some progress on this last night and finally came up with a plan for what I actually want to achieve with this, which has certainly taken long enough. I realise that after all is said and done, I've always had an affinity for the 90's Laney's so I'm working from that 4+CF Laney topology which I think I prefer to the SLO 4+CF overall. I'm going to modernise aspects of the design slightly.
That is my starting point. The cool thing about this 3 tube topology is that since I honestly will never use a clean channel (I have a superbass for that itch after all) I can use the 6th triode as tube send buffer and then implement a simple LND150 return stage for a full functioning loop with minimal fucking around and thus allowing me to use the amp as a poweramp. This leads to some interesting choices of how to split the filter nodes up which I haven't quite consolidated in my mind. If I were trying to minimise the number of nodes I could run the send and return stages from the PI node no bother; stages 1 & 2 from one node and then 3, 4 & CF from another. That breaks from the old Laney & SLO topology and should still be stable as there will be no in-phase stages to cause oscillation at the filter cap. The IRT does something similar and even has less filtering than that, where the last 2+CF are tapped from the PI node! Less filtering results in less voltage drop along the preamp. I'm less concerned about noise as I'm using a regulated screen supply which feeds the PI node.
I'm hoping to spend a few hours on sunday working on this
Loop wrote:I’m currently shopping for a 1996 Red Dodge Viper with yellow wheels. Who gives a shit about taste?!
Minor update. I've not done any work on this is a while as I'm in project set up hell at work. I've been mulling over the best way to do the FX return and PI for this for a while and it kinda coincides with some ideas for a mini ecc99 power amp project I've been kicking about for a while so I have merged the two into a PCB as an excuse to learn how to design PCBs. Thus I give you:
It's a very compact board (100x55) and features an LND150 input buffer, LTPI with presence and depth options and also options for a fixed or cathode biased ecc99 power amp. In this build, I'll be skipping the ecc99 and using it to feed big bottles. I've designed this around using a ZVS driver for the B+ so a very cheap 4W power amp can be built. Once I have verified the layout and done a bot of tidying I will order some boards. I'll probably add a few extra ground points as they always come in handy.
Loop wrote:I’m currently shopping for a 1996 Red Dodge Viper with yellow wheels. Who gives a shit about taste?!
I’d love to see an updated layout when you get it sorted. I’ve been toying with the idea of regulating the screens on a Fender’ish build.
