 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
2.99.161.121
What to do with a 5K SE OPT into 5 ohms? LL1682.It's pretty much perfect for 45 tubes, but spending that kind of money and using 3 DHT stages isn't what I had in mind. Priorities, or what I tell myself are my priorities, are:
1. All DHT. I have to be dragged kicking and screaming into using IDHTs
2. Prefer no cathode bypass caps. And not going to make a sandwich of 23 esoteric types either.
3. 2 stages. I don't use a preamp, but 2v in from the DAC.I normally use 4P1L > PSE 4P1L. Totally satisfied with that, and I could build an amp which was 4P1L into a single 4P1L. But that's more like 3.5K into 8 ohms. The LL1682 would be about 8K into 8 ohms.
Wondered about 46 or 10Y. Have plenty of both. 46 won't give much output and 10Y could do with 15k OPT. And both have less mu, so less good for 2 stages.
Ideas?
Edits: 09/03/15Follow Ups:
Decided on 4P1L > a single 4P1L in the end. It's pretty much the same output as a 45. I know the sound from my PSE 4P1L amp. I have a friend with 110db sensitive horn speakers. See how that works out.
I can come back to it and change the finals at a future date if necessary. All my amps are modular and have a number of interchangeable top plates.
Thanks for the ideas, guys!
Single 4P1L into that load will have around a watt . Can you live with that ? I run 4P1L with 97dB sensitive drivers and think it's borderline , the chocolate sauce effect definitely creeps in at louder volumes .
I think the best use of those OPTX would be for a 4P1L output stage for a TPL150H AMT HF driver . 100dB sensitive , flat 5 ohm impedence
Al
10Y PSE?
I use 801a PSE (841 -filament bias, CCS loaded-, C coupled to 10Y CF -CCS loaded-, 801a PSE -5k:8 transformer-, 620V B+). The gain to output tubes grid is about 27dB.
How about a 4P1L driving a 45 through a step-up interstage. Perhaps a 1:2 or 1:2.5 step-up would suffice. A typical 45 circuit would need about 55V max swing one way. A 2V RMS input will translate to a 56V peak each way (assuming a stage gain of 10 for the 4P1L) through a 1:2 interstage. Although not a big fan of step-ups as drivers, the 4P1L has sufficiently low impedance at ~1K Ohm to handle it. Additionally, IMO a 4x increase in Zout is still low enough at 4K for driving a 45 pretty well.
Edits: 09/03/15 09/03/15 09/03/15
A step-up transformer is one good idea. I hadn't thought of that. I use LL1660 in 1:1 at the moment. I can look at different wiring options. This may be cleaner than 3 stages if it works. thanks for that. Opens up the idea of finals with lower mu. the 4P1L in PSE has a gain of 10, which is one of its virtues. The usual suspects like 2a3 and 300b and 45 and we're down to around 3.5. The 10Y and 46 have a bit more gain so they come into the picture.
Just a short while ago I was thinking about this same thing, basically how to build a two-stage all DHT amp. I wanted to drive a 2A3 and found out that with the exception of the expensive EML 20A & 30A tubes, all other small DHT's that have a low enough plate resistance to act as drivers, are also low in gain. So, I came to this step-up transformer idea using the 4P1L which has just enough gain to barely make the concept workable.The counterargument here could be that the savings from having to buy the expensive EML tubes are put back in towards the intersrtage, so there's really no net savings. At least (and this is besides other possible advantages of transformer drive), the transformers will last indefinitely and not deteriorate as tubes do and you could also re-use the transformers in other circuits down the road. In the following link is what I had in mind for a step-up transformer with a 1:1.8 ratio (I'd use connection '7' as seen by scrolling a bit down the page).
http://softone.a.la9.jp/english/RC20/RC20.htm
Edits: 09/04/15 09/04/15 09/04/15 09/04/15 09/04/15
~!
The Mind has No Firewall~ U.S. Army War College.
Thank you for bringing up this tube as another option. However, even though it could be used as a driver, I don't think it would be best suited to that role. Looking at the RCA spec sheet, this tube has an Rp upwards of 60K at the plate supply voltages that we'll be seeing as typical in our case, those being around 400~500V. At these B+ levels, the OP of the 841 tube will fall at somewhere around the 200~300V range. This places it at the lower and non-linear portion of the plate curves which means higher distortion. We could use a choke load to bring the OP higher, to say the 500V range where both the plate resistance (at 40K) and linearity are better, but then we'd need a special choke. IMO, an Rp of 40K is still high for a choke, but it could be an option.
However, the bigger problem with this tube as I see it, is its high Miller C. With a figure of approx. 250pF (also taking into account some parasitic capacitance) it would present a challenge to whatever is ahead of it. Additionally, if we need good high frequency extension and phase linearity, the input grid resistor will need to be quite low. For example, if we were to design for a high end roll off with a -3dB point at 40KHz, it would require a grid resistor in the 15K Ohm range. This would obviously load down the preceding stage and along with the high Miller C only low Zout sources will need apply. This will in many cases necessitate the use of a preamp, something that many of us would like to avoid.
841 is super tube as VAS (I love it, and use it- I have about two dozens-).... if you comply with a few rules:
- use DC filament supply, 841 very sensitive to filament hum;
- I prefer filament bias-;
- Use large B+ for lowest distortion (sometimes need 200-250Vpp swing);
- I prefer lower dissipation operating points (about 400-500V, below 10mA);
- use anode load with extra large impedance (extreme large anode choke -I use 600H-, or appropriate CCS -for example DN2540 - IXTP01N100 cascode-);
- use if possible the greatest load after the 841 -IMHO datasheet suggested 250k is on the lower side-. I think, that another stage (CF or driver) is necessary.
If you keep this recommendations, it's not so difficult to build great 841 based first stage.
I suspect the graph you show is of a 841 loaded with a CCS. Anode chokes are pretty much out of the question unless you accept a sincere penalty in bandwidth.
dave
"I suspect the graph you show is of a 841 loaded with a CCS."
Yes, it is my 801a PSE first stage transfer graph (10Y CF as load).
Choke loaded graph much worse (due to the choke own capacitance) about 12Hz-25kHz (-3dB).
Why not an input step up transformer? A Lundahl ll 7903 in 1:4 would be a good solution, provided of course that your DAC has a low output impedance.
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
