audio-talk

Ray, could you not build the MoFo as per standard, and do any frequency sculduggery externally ?

That would make the amp more versatile, and perhaps easier to scullduggerise for other purposes ?

pre65 wrote: ↑Sat Dec 26, 2020 3:20 pm Ray, could you not build the MoFo as per standard, and do any frequency sculduggery externally ?

That would make the amp more versatile, and perhaps easier to scullduggerise for other purposes ?

Yes, of course I could Phil, but that's not what I want to achieve. My aim is to build a dedicated amp for a full-range drive unit in an open baffle build and to bandwidth limit it specifically in an attempt to eliminate external skullduggery.

I have several other amplifier options that are versatile.

simon wrote: ↑Sat Dec 26, 2020 2:55 pm Hmmm, yes, I suppose so. I'd never really thought of it in that way. Might need to read MJ again to understand how the two filters act together - I've only ever sized them in isolation.

I probably ought to get hold of MJ's books, though I'm not sure how much I'll understand!

Of course, even built as per the original design you have multiple filters, and don't forget the one on the output too, we just want to change the corner point at the bottom end. I say 'we' Simon as we seem to be heading in the same direction?

Yes, quite possibly Ray, if I pursue MoFo. I want to couple the amps directly to the speaker driver without a crossover. Perhaps one of Nick's dogmas but what I've heard doing this I liked. I might end up 3 or 4 way so I'm thinking about mechanical roll off of baffles as well as electrical roll off within the amps.

MJ's books are very good, but the design one isn't an easy read if you're not technical. Well, actually, it's very well written, but there's a lot of theory, inevitably, in it.

As I see it R1 and C1 is a filter, as is L1 and C2 (and R5). Sizing each one is fair enough but I need to think about the implications of their interaction now.

I would ignore R1 I would look at the filter with R3+P1 also R5 will be swamped by the loudspeaker load

Right you are Nick

Nick wrote: ↑Sat Dec 26, 2020 7:24 pm I would ignore R1 I would look at the filter with R3+P1 also R5 will be swamped by the loudspeaker load

Thanks Nick. In calculating a corner point for C1 a few posts back I used R3 + P1. I assumed 50% of P1 (R3+P1=60K) in calculating a value of 15nF for C1 to give a corner point of 180Hz. Retaining a value of 15nF, that means the corner point will lie somewhere between 151Hz (=70K) and 212Hz (=50K) depending on where P1 is adjusted too. I guess the adjustment of P1 will tend towards the centre of the available adjustment but something to consider.

I'm tempted to just buy some 15nF caps and Triad C-56U inductors and give it a try - I think my 300B headphone amp will have enough output to drive the MoFo directly until I build the 2P29L stage. I'm assuming the MoFo will be biased at 1.7A.

Small caps are pretty cheap Ray, might be worth buying a few smaller values too so when you bias it up you can adjust on test and add capacitance till you get the right value for the corner, then replace with a nice one?

Well, to be pedantic, remember that both the +vcc and gnd is considered ground at AC, so the effect of P1 is ( the lower leg ) in parallel with ( the upper leg + 10k ).

Not that it affects your choice, but just saying.

I counted up and I have a spare PCB leftover from the group buy so I'm just ordering a set of 'plain jane' parts to do a 'test' build with the smaller cap and triad inductor. I've also dug out an old 19.5V laptop brick and it tests OK so that's the power supply taken care of. I'm guessing it'll be after the new year before I get my hands on the parts though.

Will be seeking lessons in measuring the frequency response in due course...

As I mentioned before, I plan to use the Bartola 2P29L pre-amp to drive the MoFo buffer.In Ale's blog on the circuit he uses Rod Coleman DHT filament supplies;

http://www.bartola.co.uk/valves/2017/05 ... ht-preamp/

but I have a pair of Andrew's DHT supplies that I would like to utilise as I understand they operate on the same principle - I believed I just need build them for 9.2VDC output and slot them into the 2P29L schematic in place of the Coleman reg, however, the 120mA notation on the schematic has thrown me, compounded my Mr. Ohm telling me that with 7.1V across the 51R filament bias resistor it will pass 139mA. With both the Coleman and Andrew's DHT supplies you adjust the voltage only. Am I over-complicating this and I do just need to set the 9.2V output from the DHT supply?

http://www.lyrima.co.uk/dhtreg/dhtRegIntro.html

Surely the important voltage is that across the filament ?

Yep, the numbers don;t make sense, The valve wants 2,2v @120ma. Andrews boards will regulate to a voltage not a current, so may not be idea for filament bias. The Coleman regs are just set for a current, which is ok for this, but makes them less useful for normal heating. If its 51R resistor than as you say, it will develop 6.12v, so with the 2.2v of the filament, Andrews supply would need to be built for 8.32v.

TBH, with that circuit, I would just make a CCS either using a depletion mosfet or a LM317.

Thanks Nick, it's reassuring that I've at least learned some basics!

Why would you use a current sink instead of the filament resistor given the gyrator on the anode? As it happens, I have a spare pair of the K&K Audio MosFET-based constant current sinks. I presume you mean the same arrangement as in my 300B SE-OTL;

Ray, re the numbers not making sense. As well as the filament voltage/current there is the cathode bias voltage and current to take account of.