Bottlehead Forum
General Category => General Discussion => Topic started by: Deke609 on July 01, 2019, 08:12:18 AM
-
I think I finally have a tenuous grasp of how a basic grounded cathode typology SET amp works, and how to determine/choose the various output tube values (anode load, cathode resistor, operating point, etc.). But as stated, my understanding is only tenuous, and for it to really sink in (or to figure out if I still have it wrong), I need to design and build a super simple and cheap SET amp.
It doesn't need to sound good or even passable. I don't really care about bandwidth or saturation at higher outputs, so the OTs can be pretty small (and cheap). Nor do I care about how much power it puts out -- I just need to be able to connect it to a cheap speaker or pair of headphones to check that it is in fact amplifying the music signal. It can be DHT or IDHT - doesn't matter.
All I care about is that what I'll call the "amplification circuit" -- namely, the driver stage and the output stage -- works reasonably close to expected. I'll use PSUD to figure out a simple power supply/rectification circuit.
I want it to be fairly cheap to build -- say $200 max. But even cheaper would be better. I figure I can get a cheap toroid power transformer and some cheap Hammond OTs (or re-purpose the OTs from the Nickel Wonder pre-amp).
Where I'd like some help is with the selection of tubes. Can someone suggest a suitable (i.e., cheap) driver tube and output tube combination? (I do have a bunch of iffy 45 tubes, and 2 pairs each of the stock 2A3s ans 12AT7s that came with the SII kits; but figured something lower-powered might be cheaper to build - but I will use them if recommended).
I realize that poring over tube datasheets to find suitable tubes is part of the scratch design process, but my focus for this project is simply getting driver and output tubes to work together, not so much with designing an amp from nose to tail. And at this point my tenuous grasp of what I'm trying to do is using up all my free neurons, and i fear trying to cram any more new info into my head will brick my brain.
Many thanks in advance,
Derek
-
The Edcor GXSE output transformers are a decent step up from the small Hammond universal stuff. Depending on what you're doing, you could also use our SEX iron upgrade package.
I would suggest starting with a 12AU7 driving either a triode strapped EL34 or triode strapped 6V6.
-
Many thanks again PB.
Some quick follow-up questions about the S.E.X. 2.1 iron upgrade kit:
(1) Does it have the same values as the SII kit, 20/40H choke and 4K OT?
(2) Could the OT be used for conventional, series feed? Probably not optimally, but would running a lot of DC through it damage it?
The possibility of using this iron is attractive, particularly if the OT can shrug off series use. This would allow for a really scalable project: stage 1 - series feed; stage 2 - parafeed with resistor loaded anode; stage 3 - parafeed, but choke loaded; stage 4 - ditto, but with C4S. That would be most cool.
many thanks,
Derek
-
Yes, it's a pair of 20H/40H chokes and a transformer that can do duty at 4K or 8K. These are parallel feed parts though, there's no airgap in the output transformer.
-
Thanks PB.
These are parallel feed parts though, there's no airgap in the output transformer.
Right, so what I was wondering is whether, notwithstanding that they are designed for parallel feed and not air-gapped, the OTs can be put to service in a series-feed circuit -- or would they not perform at all, or worse, be wrecked by the high voltage DC?
many thanks,
Derek
-
The transformers are not designed for DC current.
-
Thanks Doc. I did a bit of reading and my take away is that the DC will saturate the core and possibly lead to overheating and destruction of the windings. In short: too risky.
So I may just start with parafeed.
cheers,
Derek
-
That's not really the problem. For a series feed output transformer to tolerate the standing DC current flowing through the primary winding, the laminations are all stacked on top of each other, so that all the E laminations are stacked on top of each other and all of the I laminations are stacked on top of each other. They are butted together and a piece of thin material is placed between them. This is the air gap that you will read about, and without it the DC current will saturate the transformer. A well saturated transformer will perform poorly.
-
Thanks PB. So if I Understand you correctly, you're saying that damage to the OT windings is unlikely from series use, but it will sound like hell -- which is what I expected from what I've read about air-gapped versus non-air-gapped.
I've ordered the S.E.X. iron upgrade since I am ultimately interested in parafeed amps and there are not a lot of sources of parafeed designed OTs. I will just skip doing series feed to be extra safe. From what I can see, there's no special learning to had there that I won't get from doing parallel feed, other than seeing the transformer do double duty in series feed.
thanks again,
Derek
-
Yeah, a parafeed transformer with DC running through it will have miserable power handling. You would have to derate it so much in series feed to get good bass extension that it wouldn't be worth using that way. I'd suggest playing more with the front end, trying different input tube strategies and types. You might want to play with direct coupling vs cap coupling, resistor vs. choke vs. active plate loading, pentode vs. triode. Tube vs. SS rectifiers and L input vs C input power supplies are other things to play with.
-
The iron upgrade kit is also a bit more flexible since the plate choke has a high and a low current configuration. For a series feed output transformer, a larger air gap will allow for more DC current, but with a loss of inductance as well (poorer LF response). The OT manufacturer has to figure out where to best set the gap to please the most people, or they have to deal with the negative consequences of putting a tap on the primary and having a higher current/lower impedance configuration with part of the primary winding not connected to anything.
From my limited experience in all of this, unused copper on an output transformer is something best avoided! For a plate choke, it's far less insidious.
To do an EL34 triode strapped with the iron upgrade kit, you'll want to start drawing 4K load lines on the plate curves. You might also look at the 6Y6GA/GT. They scare people off with the low plate voltage rating, but I'm sure you could run them at 250V P-K and 60mA as a fun experiment, then try the 4K and 8K loading. You could also reduce the plate current and try the entire choke winding and both impedance configurations again. They are inexpensive enough that you don't have to care if you burn a few up.
-
Thanks Doc and PB. All excellent suggestions. I'll be starting with output tube loading experiments: resistor --> choke --> choke and C4S. Interstage direct coupling sounds cool and challenging, so maybe that would come next.
@PB: following your earlier suggestion, I'll be starting with triode-strapped EL34s. Do you have a recommended way of strapping them? Suppressor-to-cathode and screen-to-anode? Or both to anode? There appear to be multiple strapping methods and I'm not that interested right now in experimenting with different ways of making pentodes behave as triodes - I just want a strapping method that results in a tube that I can treat as a triode without additional fuss.
many thanks,
Derek
-
Yes, suppressor to cathode and screen to anode.
-
Many thanks
-
@PB - maybe my Google-fu is particularly weak, but the only plate curves I can find for a triode strapped EL34 is gsmok's (of DiyAudio) "Triode Loadline Simulator": http://www.trioda.com/tools/triode.html (http://www.trioda.com/tools/triode.html)
It looks like an awesome tool, assuming it is reasonably accurate. OK to use, or should I keep searching?
many thanks,
Derek
-
Mullard has triode curves:
http://www.mif.pg.gda.pl/homepages/frank/sheets/129/e/EL34.pdf
See page 9.
-
Fantastic. Many thanks PJ.
Having already reviewed a set of Mullard data sheets earlier today, I felt a bit embarrassed when I checked your pg. 9 and saw the curves. I didn't understand how I could have missed that. But rechecking the Mullard sheets that I had found, there are no curves for the tube in triode mode, just a table of triode operating points and a triode "performance" graph that appears to plot current, voltage and distortion against power output. Lesson learned!
cheers,
Derek
-
Since you have a Beepre, you've got some voltage available for the next amp in line.
For absolute sound quality with simple design, you could try a single stage/tube amp.
Something like "pieters take on a DHT headphone amp". BH also used these tubes with good succes in Smash.
A 4K load would fit with these.
-
The Mullard datasheet should have them.
-
https://frank.pocnet.net/sheets/129/e/EL34.pdf
-
My original search may have turned up the only Mullard sheets that don't. From the National Valve Museum: http://www.r-type.org/pdfs/el34.pdf (http://www.r-type.org/pdfs/el34.pdf)
But the sheets PJ linked to and another Mullard version I found this am do have the triode curves. So all good.
cheers and thanks,
Derek
-
@Tom-s: thanks. Yeah, that's the one I found this am. I had looked at Frank's EL34 page last night, including the Phillips and Svetlana sheets and a couple of others, but having already reviewed the Mullard sheets from the National Valve Museum (which as best as I can see does not have triode plate curves) I wrongly assumed that all Mullard sheets would be the same. Not so. Lesson learned: different versions/datecodes have different data.
cheers,
Derek