Johan: I was under the impression that when using a lower than optimum load increases output power and a higher load decreases output power. Is that not always correct?
"Taj/TubeMack: For the 6GK5 with a 10M45S CCS, check out Poinz' schematic in posts #6 & #7 here:"
Not really interested in changing topology, but would have concidered a "Drop in" replacement for the EF86 with just some componet value changes, removing the pentode parts of the driver circuitry, and perhaps that 10M45S CCS!
Parb, and Bipjppop, Welcome aboard!
Not really interested in changing topology, but would have concidered a "Drop in" replacement for the EF86 with just some componet value changes, removing the pentode parts of the driver circuitry, and perhaps that 10M45S CCS!
Parb, and Bipjppop, Welcome aboard!
Last edited:
I think if I'm going to draw this up, someone had better give me a scratch version to follow. Not much of this insane-voltage stuff makes much sense to an old SS guy (at least in terms of execution, anyway). I wouldn't know how to wire up a10M45S CSS. (In fact I'm off now to Google this part number).
..Todd
The 10M45S data sheet is here.
TubeMack: I wasn't suggesting Poinz' topology (it won't work for 50W anyway), just referenced it for the 6GK5 and it's 10M45 CCS. I am hoping that it's more-or-less a drop in with some new associated component values....like the 240V B+ becomes 320V across the 41K anode load R shown in Poinz' schematic and adjust the 10M45S for 9 ma or so (which may require some other R values)...but, again, I'm still a newbie on the tube side of the schematic. The power supply stuff makes perfect sense to me.
The anode resistor is replaced by a 10M45S/current set resistor combo, the RC bias network is replaced by a red LED, and (naturally) the screen grid stuff goes.
The EF86 is good sounding, but so is the 6GK5. The resistance to slew limiting obtained from high gm tips (IMO) the scale in favor of the 6GK5.
It pays to remember that Mullard published a number of designs to stir up demand for tubes they made. A topology can be used with other types, which may be better suited to the job at hand. By all accounts, the Harman/Kardon Citation V is a nice amp. Look at the schematic. It's Mullard topology.
That "Sams" drawing is flawed, but it still gets the point across. Look at Jim McShane's site for the correction.
Thanks Eli! Sounds great to me......now I'm getting tempted to build this....I have a pair of Triode Electronics PA-060 power transformers and a pair of Hashimoto HW-60-5 (60W 5K SG) OT's looking for a project...not to mention a small pile of 6GK5's.....
The 60 W. O/P trafos lack the heft needed for use with UL KT88s and GNFB. Their cores will saturate from the LF error correction signal.
You can use the "iron" in a Mullard circuit, if the "finals" are triode wired. Another option is to use "exolinear" style circuitry, as favored by Doug Piccard (Bandersnatch). That style does not incorporate the O/P trafo into the NFB loop. You'd wind up with full pentode mode and regulated g2 B+. Hashimoto O/P trafos have a good reputation and an "exolinear" setup could work out well.
The 60 W. O/P trafos lack the heft needed for use with UL KT88s and GNFB. Their cores will saturate from the LF error correction signal.
I guess it's back to trying the Hashimotos in a KT88 version of Poinz' EL34 MM.....
Would triode wiring the finals in this circuit be sub-optimal? The B+ is about right according to TDSL....
Last edited:
I don't know if a 6GK5 pair LTP produces the gain needed to drive KT88s. I've uploaded a CRUDE schematic of "exolinear" topology. You'd have to use cascodes in the LTP, but the high gm of the 6GK5 makes it an excellent choice at the bottom of a cascode. Perhaps "Bandersnatch" will post in this thread. He has experience in constructing hybrid cascodes and such an "animal" might be best.
Attachments
I think it's time to request a quote from electraprint for the OT's. I don't think it will hurt to see just how much more they will run over the edcors. i'd love me some Hashimoto's, but the 100w units would really thin the wallet. Unless someone has a line on some?
The specs i'll send for the EP OT's will be 100W. 40% tap, 4K
The specs i'll send for the EP OT's will be 100W. 40% tap, 4K
Thanks Eli! unconventional use of the screen taps..........hopefully bandersnatch will chime in here..
Back to the existing topology of this thread.....how much power out is expected at 460V B+ and a reasonable bias current?
Last edited:
Yeah, it's a "sneaky" way of implementing short loop plate to grid NFB. You'd use regulated g2 B+ in pentode mode and take your Hashimotos to their limit.
60 W. is a very reasonable power expectation.
6fq7
"and mentions going with a 12UA7 for some strange reason. I don't know whats up with that?"
Years ago, Jim McShane advised me to use 6FQ7/6CG7 in place of 12AU7 and I have used this advice with outstanding results time and again. The 6FQ7 is electrically identical to the venerable 6SN7.
"and mentions going with a 12UA7 for some strange reason. I don't know whats up with that?"
Years ago, Jim McShane advised me to use 6FQ7/6CG7 in place of 12AU7 and I have used this advice with outstanding results time and again. The 6FQ7 is electrically identical to the venerable 6SN7.
hey-Hey!!!,
No need to thin the wallet to far. The Hammond 1650R is a 5k/100W unit that works very well. So well in fact I abandoned an expensive project based on reproductions of the Peerless S-271-S I undertook. The Hammond is that good...
On current, you'll get a maximum listed by the manufacturer, that is the DC idle current the tubes are passing with no signal. You'll be well under this, even for KT88's.
cheers,
Douglas
hey-Hey!!!,
No need to thin the wallet to far. The Hammond 1650R is a 5k/100W unit that works very well. So well in fact I abandoned an expensive project based on reproductions of the Peerless S-271-S I undertook. The Hammond is that good...
On current, you'll get a maximum listed by the manufacturer, that is the DC idle current the tubes are passing with no signal. You'll be well under this, even for KT88's.
cheers,
Douglas
Why I ask is because electraprint requires me to list this spec. Are you saying they want to know the max quiescent current? How high above this do I spec? Whats the best guess for this figure?
TubeMack: Here are some specs for off-the-shelf transformers; a few data points to suggest that you want around 300ma for the electroprints based on the following:
Hammond 1650R is 5K & 318ma (100W)
James JS-6295HF is 5K & 250 ma (100W)
Hashimoto HW-100-5 is 5K & 300 ma (100W)
I'm unclear if the 4K secondary would require raising the ma rating.....
I imagine Jack at Electraprint would have advice if you told him that you are doing UL KT88 PP with 460 B+..
Looks like Edcor's CXPP series max out at 60 watts off the shelf, and I could not find the current rating for those....
Hammond 1650R is 5K & 318ma (100W)
James JS-6295HF is 5K & 250 ma (100W)
Hashimoto HW-100-5 is 5K & 300 ma (100W)
I'm unclear if the 4K secondary would require raising the ma rating.....
I imagine Jack at Electraprint would have advice if you told him that you are doing UL KT88 PP with 460 B+..
Looks like Edcor's CXPP series max out at 60 watts off the shelf, and I could not find the current rating for those....
Last edited:
.
Look again. Edcor offers a number of CXPP100-* (100 W.) models. I suggest that the CXPP100-MS-3.3K be used in this project.
Examine the primary impedance issue from a perspective of the voltage swing available from the tubes and the step down ratio of the "iron". A 3.3 KOhm primary impedance fits the 60 W. target well. The voltage swing available from KT88s is only slightly greater than that available from EL34s, but nearly 2X the power is available. In order to pass that power to the speakers, the similar voltage swing has be stepped down less. An end to end primary impedance of approx. 4 KOhms works for EL34s and the "35" W. they produce, but it steps down too much for "60" W. KT88s.
BTW, Edcor "iron" is not just for cash strapped builders. Look here and you'll see that Edcor's stuff is quite good and a superb value.
Like this? Removed the end bells, bead blasted the blue (to remove the gloss), then sprayed with Rustoleum satin black. Finishing touches were stainless screws, with stainless acorn nuts. Kinda looks like an electraprint now..
