power supply for a big bass amp-> sound ok?

[black_labb]

Has there been any progress on this? I'm currently building something similar, so I'm interested to read how you're going.

If you've not yet purchased an OT I'd suggest investigating the Hammond 1650R. Although rated at 100W it has been used successfully to put out 200W with a B+ of 750V - http://www.dogstar.dantimax.dk/tubestuf/dzart-4.htm

If you're after one in Aus send me a PM and I'll tell you how I got mine for a little cheaper

no progress, just research. i'm planning on ordering some of those torroidial trannies in the next week (the big one, and a smaller one for a 30-40w guitar amp that i have the output tranny for).

i'm thinking i might do a hiwatt style tone stack, presence control and PI. the front end will be somewhat different though. i'll probably end up using some 6bl8's in there too.

I'm also considering of building a 6xKT88 amp and was wondering how much the plates supply filtering should be. According to the schematics of different famous bass amps it goes from 50uF (2x100uF caps in series) in Ampeg SVT all the way to 1100uF (10x220uF in parallel and series) in Mesa Bass400?

i'd probably say that the higher uf levels would be better. the stiffer power supply would help keep those transients clean and punchy. im going to aim for 220-330uf i think.

from your description i'd say that the boogie it is 550uf, not 1100 (caps in series drop the uf, in parrrallel add, opposite to resistors)

[andrew_k]

no progress, just research. i'm planning on ordering some of those torroidial trannies in the next week

I put in an order for their 360-0-360 tranny for my twin 6550/KT88 amp this morning. Modeled the PS w/ bridge recto, LC filter (50uF), 615v RMS on the plates and 397 on the screens. You might not need the big 500v CT depending on your rectifier choice.

[black_labb]

are you sure you put the right values in? thats jsut a simple bridge rectifier with no fancy bits or multipliers?

360 * ~1.41 should give the height of the peaks, which comes to ~510. you wouldnt eb getting any more voltage than the peaks without increasing the voltage from what it is now. what progam are you using? for reference, the 500vac tranny should give me roughly 700v, using a full wave rectifier (or the same with a bridge rectifier, but using the full wave makes sure that i use both windings.

should of mentioned that you were ordering from there, we could of tried to save some money on shipping and put out orders together. oh well.

[black_labb]

are you running it as 720v into the bridge rectifier with a choke input? that could be the right level.

im running the choke input for grids and the preamp, but the plates will be run off a separate FW rectifier to get the 700v.

[GainFreak]

rom your description i'd say that the boogie it is 550uf, not 1100

You're right. It's 1100+1100 in series=550uF.

[Steve Conner]

I don't think there is any firm answer as to what is the "right" amount of plate supply filtering. Newer amp designs use more, but that's probably just because electrolytics have got smaller and cheaper for a given uF/voltage over the years.

Choke input filters always struck me as a bad idea nowadays. For optimal performance, you need a swinging choke, and nobody makes those any more. The choke also radiates a lot of hum, far more so than if it were used in a normal CLC filter.

Choke input filters were really only useful back when the smoothing caps were all huge paper in oil things, and you had to minimize the space they took up, and you had tube rectifiers that couldn't tolerate peaky current. Now that you can get cheap electrolytics, and silicon diodes that don't wear out from high peak currents, there's really no reason to use one.

In other words, you've probably ordered the wrong transformer.

[andrew_k]

are you sure you put the right values in? thats jsut a simple bridge rectifier with no fancy bits or multipliers?

I'll be running both HT secondaries in series (no grounding), with a bridge rectifier and RC filtering. A choke capable of feeding the plates was too expensive for me. I've revised my order to a 325-0-325 PT as I'm actually getting TOO MUCH voltage on the PSUDII model. I will be starting another thread to ask for some help on that once I'm more confident in my existing model.

I have not yet received an invoice nor paid for my order so there's still time to combine if it will save us some money. I'm in Melbourne, btw. John from AnTek was very quick to answer my initial questions, perhaps send him an email about combined shipping.

[black_labb]

makes sense, but 650vac is still a huge amount, kt88's are only able to handle 800vdc on the plates max, and unless you do a choke input (which you dont want to do) i doubt you will be under 900v. i'd suggest the 450v tranny if you want ~625v. a 225-0-225 would work as well if you want it in series.

[andrew_k]

!

You're absolutely right. This is my first time using PSUD and I made a big mistake by putting in current sinks for maximum draw at each stage. What I failed to consider is that the power section will never have a chance to use maximum draw after the initial inrush of 900V DC wipes out the valves
I will use a 230-0-230 to make ~640DC. Thanks for steering me in the right direction before it was too late!

[Steve Conner]

Yep, I think you want something like a single 500V secondary for use with a bridge rectifier. So if you were getting one of those Antek toroids, you'd want two 250V windings to hook in series, or two 500s to parallel.

You could also make the two 250V windings independent, each with its own bridge rectifier and smoothing cap, and then stack the resulting two 350V DC supplies in series. Then you can draw 700V from the top of the stack for your plates, and 350 from halfway up for your screens/preamp.

In the 300PS, Fender used a single 265V (iirc?) winding that drove a full-wave voltage doubler arrangement. That provided 360V for the screens and 720V for the plates. The only drawback is that the 360 node is essentially fed by a halfwave rectifier. So drawing too much current from it might saturate a toroidal transformer and make it buzz. The 300PS used a huge E-I transformer that looked like it was stolen from a substation, and I doubt it would have been bothered by a little DC.

[andrew_k]

You could also make the two 250V windings independent, each with its own bridge rectifier and smoothing cap, and then stack the resulting two 350V DC supplies in series. Then you can draw 700V from the top of the stack for your plates, and 350 from halfway up for your screens/preamp.

That's a great idea, thanks! AnTek don't offer a 250-0-250, but the same theory applied to the 230-0-230 provides good voltages for running the 6550C I already have. I have more questions but I'll start another thread as this is derailing black_labb's KT88 bass amp thread.

[black_labb]

for the choke input, when it is rated at a certain voltage dc, would i base that on the voltage at the output, or the input of the choke. im thinking that by putting the input wire (with the large peaks) at the end that is better insulated from the frame, i should be able to use a 500v rated choke, as opposed to a 700v rated choke, which would be easier to source/cheaper

[Steve Conner]

The DC voltage rating of a choke is really more of a safety/legal thing, as far as I know. It's just whatever voltage the manufacturing process has been UL certified at. I've used little chokes rated at 400V with something like 2000V AC across them for short periods.

For a choke input filter, the rating you need to worry about is the choke's saturation current, which should ideally be twice the current you plan to draw. This also indirectly specifies the AC voltage rating of the choke before saturation.

If you don't know what the saturation current is, there's another reason not to use a choke input filter. However, if you try it anyway, you may get some swinging choke action going on and it might all work out.

[loudthud]

Shoulda popped in earlier on this thread but a 400VA transformer is not a good choice for a 300W amp. It wouldn't take much to overheat it. 600VA would be would be a good place to start.

[black_labb]

you texans, always need things to be big


the 400va seems reasonable to me, as it should be 800ma of current . the datasheet for the tranny has it putting out 480v at 1A, so i think it should be up to the challenge. its not quite ideal, but its better than most manufacturers would do. and 50$ for the tranny is a great deal.