| Klimon |
Please help me understand:xeye: :xeye: :xeye:
I'm planning to build a double dc darling but when redoing the designer's homework (checking bias and loadlines) the bill just won't fit.
Why do the 1626's have about 400V at their anodes? With regards to further bias-variables I'm aware of the different cathode resistor value for parallel tubes (namely half) but it still doesn't make sense. It would be a whole relief if someone could explain in a few lines how this amp is biased or what I'm not getting...
Here are links to the schematic (second one on the linked url) and the loadlines:
http://www.geocities.com/TimesSquare/1965/jeremy.html
http://www.mif.pg.gda.pl/homepages/.../049/1/1626.pdf
Many (many) thanks!!!!!
Cheers,
Simon |
|
|
| arnoldc |
The 1626 is biased at around 27V (154V - 127V).
This is a direct-coupled design and is the reason for the high B+ but the plate voltage on the 1626 is somewhere in the region of 235V (394V - 154V - 5V) considering the OPT voltage drop. |
|
|
| Klimon |
| Bless you, Arnold :angel: |
|
|
| arnoldc |
| You're most welcome :) |
|
|
| drj759 |
Klimon,
In Direct Coupled designs, the voltages of the 1626 have to be shifted to provide the proper grid bias voltage since the grid is held at 127 volts by the plate voltage of the 8532 driver stage. There is no coupling capcitor to block the DC of the first stage as there is in the first schematic. The levels also need to be shifted to maintian safe operating conditions for the 1626's to keep from exceeding the max plate to grid and max grid to cathode voltage ratings.
The 3200 ohm, 10W resistor is there to elelvate the cathode to a level to provide the proper grid bias of -27 volts (127V - 154V = -27V). The higher 394V is needed to compensate for the 154V drop across the 3200 ohm, 10W cathode resistor for the 24 ma current through each 1626 to maintain the 240 V 1626 plate voltage. The operating point for the Double DC Darling 1626's is going to be 394V - 154V = 240V at 24 ma per tube which is very close to the same operating point as the cap coupled version in the first schematic.
Don't forget to tie the heater to the cathode as shown in the DC Darling schematic to keep from exceeding the heater insulation breakdown voltage since the cathode will be at 154 volts.
Good luck. |
|
|
| Klimon |
That's been very helpfull drj, I noticed your post only now while rechecking the thread to calculate current draw (24 mA;)/tube). The tied cathode/heater totally slipped my mind...
Simon |
|
|
| drj759 |
Simon,
I’m glad the information was helpful.
Your original post was timely as I'm in the process of building up a double DC darling myself. I had just gone through the same calculations about a month prior.
I’ve built a breadboard version of the original darling and a cap coupled double darling. The next logical step is to try the DC double darling.
The driver 8532’s sounded decent but the ones I have are a bit microphonic.
My posts are still under moderation so there will be a delay before the moderators accept this post. Hopefully they will remove my posting restrictions soon (Hint ;) ).
Dave |
|
|
| Klimon |
Hi Dave,
How do the cap-coupled darling and ddarling compare? I find the doubled power (and aesthetics:o) of the 4 output tubes very attractive but also heard stories about loss of focus and parallel matched tubes quickly drifting apart.
I'm already in the phase of laying out the chassis of my double dc (will build it directly without breadboarding) but it can take another while before I've collected a decent potmeter etc. Let us know how the double RC works out and stacks up against it's siblings.
Welcome to the forum (if they'll let you in..:D)
Simon |
|
|
| drj759 |
Hello Simon,
(A bit lengthy post)
I’m a newbie to valves as of last April with the original Darling as my first effort. I’m still relatively low on the learning curve so please factor that into my comments. Also, my listening experience on valves has been limited to just my own projects. I hope someday listen to some other valve systems to compare with my efforts.
I can’t say that my efforts to date on the Darling have been a complete success for me. I blame my inefficient speakers. Resource constraints have limited my speaker choice to a pair of vintage 1970’s Large Advents that I recently re-foamed and re-capped. I probably need to use a more efficient speaker but I’m reluctant to go down the DIY speaker rabbit hole just yet.
My complaint with the Darling on the Advents is listener fatigue after about an hour of listening since I had to drive the 1626’s hard to obtain a reasonable volume level. The Double Darling with the additional output power helped to reduce the listener fatigue some but not completely.
I did notice some nice characteristics of the Darling so I want to continue experimenting with it. I thought the soundstage was very good on the Darling. I had never experienced the sensation of presence as I did with the Darling. Jazz and Pop sounded good, Classical did not fair as well as I would have liked. I was surprised how much bass I got out of the Darling with the Advents. It was a strong bass but very controlled. My Double Darling trial was done rather hastily so I’m reluctant to make any further comments on that configuration. So I want to give the DC Double Darling one more try now that I understand a bit more about valves.
Your comment about the loss of focus on the Double Darling has raised my curiosity. I’ll have to rebuild the original Darling configuration for comparison to the DC Double. |
|
|
| SY |
| The Advents just scream for a good 15-50 watt push pull amp. They were a classic match with Dyna ST-70 or MkIII back in the day. |
|
|
| Klimon |
| quote: | | I probably need to use a more efficient speaker but I’m reluctant to go down the DIY speaker rabbit hole just yet. |
A good way to get your toes wet are vintage two-way bookshelfs; they cost very little but can give excellent performance if the woofer behaves well (smooth high-end roll-off) -- just run the woofers fullrange, add a modern dome tweeter with a single motor-run cap and the result can be excellent. It's not uncommon for '70s- '80s speakers to have 90-93db efficiency. A zobel can also do little wonders (string of 6,8 ohms power resistor + 10µF cap works well on ~8 ohms woofers), your advents would probably also benefit from this when driven by darlings.
Your description of the darling sound strikes me as typical of single-ended amps (I'm an aficionado), except for the good bass-response. A very low plate resistance could account for that but it's not indicated on any 1626 datasheet I've seen.
My remark about possible backdraws of parallel output tubes is probably only valid if the tubes aren't well-matched, this is one of the only applications where tight matching really matters.
Simon |
|
|
| drj759 |
| quote: | Originally posted by Sy
The Advents just scream for a good 15-50 watt push pull amp. They were a classic match with Dyna ST-70 or MkIII back in the day.
|
Sy, I was already thinking PP but wasn't thinking as big as a ST-70 or MkIII just yet. A good point though for me to "think big" for future consideration. I've got the parts to build a version of Eli's El Cheapo sitting in the basement waiting for some attention. My goal with building the El Cheapo is to learn a bit about PP. My plan is to start with the 6AQ5's as in Eli's original design then try EL84's.
| quote: | Originally posted by Klimon
Your description of the darling sound strikes me as typical of single-ended amps (I'm an aficionado), except for the good bass-response. A very low plate resistance could account for that but it's not indicated on any 1626 datasheet I've seen.
|
Simon,
I've read a lot of comments on various message boards about the lack of bass response with SETs so I'm well aware of the issue.
I can't explain it other than the Advents themselves. The OPT iron is nothing special, Hammond 125ESE. But I can definately rattle the light fixtures in the basement with ease with a CD player directly connected to the Darling. Sends the cat running to hide. Similar but a bit stronger bass result with a plinker 6FD7 SET amp that I built.
Power supply is nothing especially exotic either, Hammond 269EX, full wave HexFreds, CRC (470uF, 100R, 470uF) using Panasonic TS-ED's for the caps (the lowest ESR that I could find in Digikey's catalog). I tried a borrowed pair of bookshelf Paradigm's and the Advents significantly outperformed the Paradigm's on the low end.
I'll keep in mind your suggestion about a Zobel. I've seen posts about Zobels and Advents but I haven't had the time to study the concept. Seems easy enough to try. I also keep an eye out for some vintage bookshelfs (I can envsion the wife's eye's rolling when I drag home more "stuff").
Dave |
|
|
| drj759 |
Sy,
My posts are still under moderation. Is there any way you can change that?
Thanks in advance,
Dave |
|
|
| pinkmouse |
| Not any more Dave. ;) |
|
|
| drj759 |
| Thanks, pinkmouse. :D |
|
|
| lndm |
| quote: | Originally posted by drj759
I've read a lot of comments on various message boards about the lack of bass response with SETs |
They don't have to be built that way and can sound incredibly good. I think that rumour must have started with mantel radios. |
|
|
| Klimon |
| quote: | | (I can envsion the wife's eye's rolling when I drag home more "stuff"). |
:D
| quote: | | I can't explain it other than the Advents themselves. |
Possibly not too low impedance at the bottom. I'm still wondering what plate resistance the 1626s have though; I once built a little 6080 set (around 500 ohms?) which had a thunderous bass performance.
| quote: | | I've seen posts about Zobels and Advents but I haven't had the time to study the concept. |
Flattens impedance (that rises at higher frequencies on woofers due to self-inductance of the voice coil) which can be handy for nearer-to-textbook XO performance or, in two-ways, attenuating the high-mids / low-treble a little (say 2kz till roll-off) = good way for fine-tuning tonal balance without degrading the sound
| quote: | | They don't have to be built that way and can sound incredibly good. |
I imagine a set with one or more of the following features would have an edge bass-wise: NFB, very low plate resistance, low impedance PS , an easy load, more output power, large bandwith :confused:...
Simon |
|
|
| lndm |
| quote: | Originally posted by Klimon
I imagine a set with one or more of the following features would have an edge bass-wise: NFB, very low plate resistance, low impedance PS , an easy load, more output power, large bandwith :confused:...
|
Is there any concievable reason why an SET should have a lack of bass?
I suggest it is the use of an undersized OPT that is not only limited in inductance and straining under the DC, but saturates more with music peaks, and hence rolls off the bass.
The right OPT can and is, these days often specified by DIYers and the plate resistance doesn't have to be low, just correct. The right load is just the right load for the job, a class A amp does not need a low impedance power supply as badly as a class B amp wants one, and NFB is not mandatory especially when the response is already dynamically flat.
Naturally your mileage may vary, but it hurts my senses to see this rumour perpetuated whilst its reason is borne from practicality and not, IMO good engineering. |
|
|
| drj759 |
| quote: | Originally posted by Klimon
I'm still wondering what plate resistance the 1626s have though; I once built a little 6080 set (around 500 ohms?) which had a thunderous bass performance.
|
Simon,
For clarification, I was not claiming thunderous bass performance like from a plate amp on a subwoofer. Just decent, controlled bass that was strong enough to rattle a couple of ceiling mounted light fixtures once in a while (and scare the cat :devilr: ).
I came up with a ballpark rp of about 2400 ohms for the 1626 by eyeing the plate curves displayed on my monitor screen (225 V at the plate, -24V on the grid and about 25ma). Doing the same rough estimating process (different Q point) for the 6080, I came up with about 800 ohms that is close to your 500 ohm approximation.
| quote: | Originally posted by Klimon
I imagine a set with one or more of the following features would have an edge bass-wise: NFB, very low plate resistance, low impedance PS , an easy load, more output power, large bandwith :confused:...
|
A few comments to add to Indm’s:
I don't view NFB or very low plate impedance as prerequisites to a successful design. I view them as tools in the design space that might be useful for certain situations. It also depends on what you want to achieve. Generally NFB will lower impedance but that alone does not guarantee success. The couple times I tried NFB on some of my SET projects (not the Darling), it sounded lousy. I probably: 1) provided a solution where I didn't need one or 2) didn't design it correctly. Certain design points might require low rp (OTL, PS Regulators and maybe your design point for the 6080 as well).
More output power is generally good but I'm not convinced I need 100W for my situation.
I try for critically damped, low impedance PS.
I view more bandwidth as being generally good if the amp remains stable. |
|
|
| Klimon |
| quote: | | Is there any concievable reason why an SET should have a lack of bass? |
That has simply been my experience with admittedly a small number of homebuilt amps; e.g. el84 triode-strapped into decent 5k opt gives less than acceptable bass-performance (no attack, sounds slow and weak) without NFB. With a small amount of plate-to-plate feedback it works very well for me. The 6ck4 I breadboarded (now being built) wasn't impressive either in the bass-departement.
No fundamentalism here, just concluding out of my experiences.
| quote: | | I came up with a ballpark rp of about 2400 ohms for the 1626 by eyeing the plate curves displayed on my monitor screen (225 V at the plate, -24V on the grid and about 25ma). Doing the same rough estimating process (different Q point) for the 6080, I came up with about 800 ohms that is close to your 500 ohm approximation. |
Nothing special then but 1626 plate resistance possibly lower than what you calculated as 6080 is actually 250. There's a 7236 set on my short list which I expect to confirm that very low plate resistance coupled to an opt with much too high Zout (4k) is one recipe for impressive low-end.
Simon |
|
|
| drj759 |
| quote: | Originally posted by Klimon
Nothing special then but 1626 plate resistance possibly lower than what you calculated as 6080 is actually 250.
|
I went back and checked my original Darling notes and I used a rp of 1700 ohms in my calculations. Seemed to work out with the voltages that I measured. I used the RCA data sheet for the 2400 ohms that I calculated yesterday. I found that the tung-sol data sheet at franks tube data site listed the 1626 rp as 2500 ohms (both 1625 and 1626 are listed on the same data sheet). I'm not sure how the data for Class C operation in the RCA data sheets changes for Class A.
http://www.mif.pg.gda.pl/homepages/.../127/1/1625.pdf
My calculation on the 6080 did seem a bit high for that low of mu. I should have known better. My haste. I probably need to be more careful for what I put forward on the web.
| quote: | Originally posted by Klimon
There's a 7236 set on my short list which I expect to confirm that very low plate resistance coupled to an opt with much too high Zout (4k) is one recipe for impressive low-end.
|
I'm not familliar with the 7236. You might want to take a look at pages 557 and 558 of RDH4. Figure 13.14 illustrates the effect on power output and distortion versus plate load for a 45 triode. Principles should be the same for other triodes.
Dave |
|
|
| drj759 |
Simon,
I finally had a chance look up the 7236. I think I understand a bit more your direction with the 7236 but I am curious how that relates to the information in RDH4 I referenced in my previous post. Again, just me trying to understand the concepts. Thanks.
Dave |
|
|
| Klimon |
| quote: | | I found that the tung-sol data sheet at franks tube data site listed the 1626 rp as 2500 ohms |
Thanks, so roughly the same as my triode-strapped el84s
| quote: | | My calculation on the 6080 did seem a bit high for that low of mu. I should have known better. My haste. I probably need to be more careful for what I put forward on the web. |
that's okay
| quote: | | but I am curious how that relates to the information in RDH4 I referenced in my previous post |
I checked the link but I'm not sure the explanation can be found in the relationship between harmonic distortion and plate load. I think the good bass of the 6080 family (with high 4-5k load like I tried, maybe also with more conventional 1-2k load) has more to do with damping factor. I'm intellectually very lazy (and more pragmatic than engineer) but will search for the technical explanation... will check back later
Simon |
|
|
|
