SE OPT for ER300b

Design and use of the various types.
Post Reply
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

Do you have a output transformer winding plan for a single-ended Elrog ER300b? If not, I would like to start a conversation regarding winding a pair of OPTs for this tube.

While compatible with most WE300b applications, the plate resistance is higher at 950 ohms and mu is lower at 3.3. Power output less than 10 watts with idle current set somewhere between 60ma and 77ma. on workable load-lines from 4700 ohms to 3200 ohms.


I am interested in an alternative to available steel C-core OPTs. I am leaning towards a nickel, dual C-core for realistic harmonic development rather than hyper-etched transients. A discussion will be helpful to narrow down the core, load-line and operating point options.


Thank you.
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

I neglected to mention the output parameters. The speaker is rated at 8 ohms, but woofer is actually 4 ohms. The impedance transition is smooth through the 800Hz crossover, but the low-frequency impedance declines from 6 ohms to 4 ohms at 20Hz. Could be averaged to 5 or 6 ohms.

The crossover is spit LF/HF and the connections are brought out for bi-wiring. Since the highest power requirements will occur in the woofer range, one possible option for the OPT is to connect the LF section to a 4-ohm tap and the HF section to the 8-ohm tap. Given your preference in for a single secondary winding in earlier forum posts I have read, I am interested in your opinion whether this is a workable use case for your winding designs.
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

Post by HSEAlark271 »

Adding to use case conditions relative to bandwidth if this helps determine whether a good fit for a transformer design.

Ideally bandwidth would be flat between 27Hz and 40kHz. This implies HF -3dB around 120kHz and LF power -3dB down to 9Hz. Interpolating curves from an omega*Lp/rp chart suggests minimum inductance to achieve the LF power cut-off of 14H for the 3.3rp load-line and approaching 16.8H for the 5rp load-line - looks a little low to me, but that is what the math indicates.

Imax ranges between 110ma and 144ma for the load-lines mentioned in A1. While expected power peaks remain in A1, I plotted headroom peaks into A2 that reach estimated Imax between 120ma and 165ma. These extraordinary peaks should occur much higher than the LF cut-off. The transformer design may limit the excursion into A2 to stay out of core saturation.


Since I do not know the magnetic properties of nickel cores, the information above is just a framework for a discussion to resolve a workable operating point and transformer design.
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Hey,

I would expect that the ER300b would work well into a 5K load. Nickel is a good choice and I can offer EI laminations.

The varying Z between the speaker loads is tough because if you try to match the load by using different taps you will alter the voltage levels that the speaker is presumably designed around.

Bandwidth on the low end isn't an issue but the -3dB point @ 120Khz to keep flat to 40Khz assumes a simple first order filter which outputs typically are not.

The design is entirely possible and we should continue the conversation to close in on some of the details.

dave
Get Your Fix
www.hifiheroin.com
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

Thanks for your reply, Dave.

I will number the detail items so we can track resolution to the final design. Unfortunately, my responses will be in multiple entries, as the forum keeps timing out before I can complete the list.

1. Core: ?

While I have tried to read all your forum and blog contributions, I am sure I have not found them all and found myself reading an Emia blog entry last night that for some reason I had missed in earlier scans of that blog. The "Permalloy" entry from June 12, 2012, indicated your preference for amorphous-cored output transformers with thoriated tungsten filament output tubes until you heard a set of 80% nickel transformers. I am curious if your opinion has held over time since that listening session.

Elrog differentiated their 300B variant by using a thoriated tungsten filament, which also caused the deviation from other 300B characteristics. I expect the use of this filament will align with your experience, but would like to confirm your recommendation for this project.

2. Laminations: ?

Whatever you work with for the selected core material will be fine.

Continued in next post.
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

3. Tube load: 5k

5k works for me. This value came out most often from various rule-of-thumb calculations. It also lands on the lower distortion side of the power output vs. distortion compromise balance.

4. Speaker impedance: ?

I was thinking of consistent tube load over the transition from 4 ohms to 8 ohms speaker impedance when I wrote the dual tap description. I was not convinced the dual tap idea would be useful without modifying the crossover.

The speakers for this project are Klipsch Cornwall III with stock drivers and crossovers. This speaker is designed to be placed against a wall to radiate into 1/4-space and the woofer height is set to reinforce a band of frequencies to compensate for the difference in sensitivity between the woofer and horns. I had not considered how the 4-ohm woofer would contribute to this sensitivity compensation. Thank you for reminding me of this element of the system design.

Based on this analysis, 8 ohms seems to be the most appropriate target output load impedance for the winding ratio rather than an averaging compromise of 4-5 ohms.

5. Passband: <10Hz - ?kHz

I was definitely asleep when I suggested a first-order high-frequency roll-off. The Cornwall has a natural falling slope frequency contour, so the objective is actually to minimize further roll-off the top octave to the extent possible in a reasonable transformer winding. It will be interesting to see what you can achieve.

Continued in next post.
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

6. Operating Point: ?

I expect there is some interaction between operating point, current swing and gap that fit an optimum operating area of the chosen transformer core. Thomas Mayer suggested operating the ER300B at 60ma. between 350v and 400v across the tube. Relative to the 5k tube load choice, plotting load lines on the ER300B plate curves suggest a design center operating point of 63ma. at 380v. 60ma. at 350v limit the output voltage swing a little more than I would like and 77ma. 390v is about the high limit I would consider. Power-line voltage is fairly stable here - typically within 5% - so 380v seems like a good design center.

7. Power output: 6.3 watts rms

A little over 3 watts is all that is required to achieve peak acoustic reference levels in our room with the Cornwall speakers. 6 watts give some headroom. I do not feel the need to push over 10 watts for my use case.

8. Gap: ?

The load lines above cross the 0-volt grid line between 100ma. and 125ma. Operating the output tube in A1 seems sufficient to meet the room acoustic levels, but I am interested in any peaking capability into A2 that can be achieved to ease clipping without compromising the impact of gapping to achieve this additional headroom over the A1 operating current swing.

I am open to your recommendations regarding optimizing these operating conditions to best utilize the selected core and size.

Continued in next post.
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

Other construction details:

9. Termination: ?

I would prefer wire leads to connect directly to other circuit components rather than terminals on the transformer, but will differ to your construction practices.

10. Mounting and cover: ?

I expect resolution of these details will fall out from choice of the core and laminations.

11. Any other details that need to be addressed?

This is the end of this pass at addressing details - 4 posts; 11 items.

Regards;
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

By number...

#1 The ER 300B does not have the same character as the Thoriated tubes I was referring to (eimac 75TL RCA 834) Nanocrystalline is an option that gives a bit more musicality than amorphous.

#2 The core material dictates the form... nickel=EI Nano and amorphous= C-core

#7 this even opens up the possiblity of 80% nickel :-)

#8 gap becomes a choice of inductance vs. current. I can provide graphs.

#9 I bring out the winding wire for termination to the circuit.

dave
Get Your Fix
www.hifiheroin.com
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

Thanks for the additional information. Let's see if we can finalize item 1 core material.

Without getting into a long discussion, I think I fall into the "musical" camp regarding reprocution system qualities. Systems with etched transients seem to dry up the sound and harmonic development seems to be lost. However, I do not like the mushy sound from days of yore. Macro-transients are important and the silence between the notes is a good indicator of balance between good harmonic support on the trailing edge and too much holdover mushiness.

While I am working backwards from speaker to source inputs and have not finalized circuit designs end-to-end, I do not expect to use an interstage or input transformer in this amplfier project. I do plan to employ a choke-input power supply architecture. Ths only reason to mention this is include any aspects of transformer cascading, or not, in the output transformer core decision.

I have insufficient experience comparing these transformer core options and do not have a fixed preference. I do know there are trade-offs in magnetic properties among these materials. I started off with some assumptions regarding C-core and nickel, but also know a complete transformer design involves many factors. Given your experience and design approach, would you nudge me in a direction towards 48% nickel, 80% nickel or Nanocrystalline?

Regards
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

I apologize for adding one more bit of input information that may impact the choice of core material. I changed from a Lyra Dorian cartridge to an EMT JS6 and found the sound fuller and more lifelike. Operated lightly loaded, the Dorian was not bad, but thinner sounding compared to the EMT. Digital source is a Chord Dave using PCM sources. So far native playback of DSD files have sounded veiled compare to PCM versions of the same recordings. I thought this might be informative relative to my perception of analytical vs. harmonically enriched playback as well as the type of sources in the current system.
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

After posting thr responses above, I reviewed older communications with other knowledgeable people focusing on nanocrystal cores to which I had not paid much attention. Two issues of interest emerged: soft bass with permalloy and deep insight into the mix with nanocrystal. Any feedback that helps fit the core choice with the horns and large woofer of the Cornwall’s along with my desire for a transparent view of the music will be appreciated.
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

If you are into the musical camp then i would take the amorphous off the table. While it has some sort of magical combo with the big powered thoriated tubes and can be affordably sized to do big power, i think it is too "clean" for the ER300B.

This leave the nano and nickel with the nickel (particularly 80%) being more organic and tuneful and the nano being "accurate" without crossing the line to analytical. For the big ER triodes I have used the nano for thomas but those are 15W outputs which is a tough job for the 49% in the sizes I have available and moving to C-cores in small quantities is horrendously expensive.

Since you only require 6W or so and need full range my natural thought would be to use 49% with the nano being second.

The bit about the lyra vs. the EMT is actually quite valuable. To me the Lyra's are all super clean and detailed leaving much of the emotion in the music on the table. The EMT's have tone and soul and tend to the musical. I would say the lyras are akin to the amorphous and the EMT's more in the nickel / nano camp.

soft bass with the permalloy has not been my experience. the bass seems to blend seamlessly with everything else. The nano may have a seemingly more impactful bass but this comes at the cost of a slightly less than homogenous sound. Almost every argument I have seen about the use of nickel in DC carrying situations has been 100% based in the technical world and is 100% the opposite of what I have experienced in the real world.

when it comes down to it I think you are at a coin flit between nano and 49%. If you are all sbout the technical and measurements then I would lean to the nano. If you want something that gets to your core with the emotion I would go for the nickel. sorry I can't be more definitive in this but at this point we are down to flavors and blending of a really good meal.

dave
Get Your Fix
www.hifiheroin.com
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

Thank you for the clear articulation of the trade-offs based on your experience. I appreciate your willingness to provide input to those of us trying to learn more.

I will go with the 49% nickel for this project.

The detail list to date:

1. Core: 49% nickel
2. Laminations: EI
3. Tube load: 5k
4. Speaker load: 8
5. Passband: <10Hz - high-frequency rolloff TBD from finished transformer design
6. Operating point: ? 63ma. at 380v
7. Power output: ~6watts
8. Gap: ?
9. Termination: wire leads
10. Mounting: ? Open frame or end bells; other options to consider?
11. Winding wire: ? Copper, copper/silver, silver
12. Saturation level: ? Any estimate of current vs. distortion or does the gap need to be determined first

Two new detail items added. I forgot to add wire material in the initial list. Assuming a signal chain up to the output transformer and good speaker cables for the Cornwalls, how much will silver secondaries or complete silver windings really add to the musucality of the 49% nickel cores?

Regards
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

The 49% nickel outputs come with brass end bells and in copper to your spec are $1400 a pair.

Silver would require a "are you sitting down" type conversation because the price goes up by multiples quickly.

The gap is determined for the best compromise of inductacne vs current at the lowest frequency of interest.

dave
Get Your Fix
www.hifiheroin.com
HSEAlark271
Posts: 12
Joined: Thu Feb 07, 2019 9:14 pm

SE OPT for ER300b

Post by HSEAlark271 »

Copper makes the most economic sense and I am not sure the improvements would be worth the cost in this project. Let’s go with copper for this order.

How do we proceed from here?
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

send me a private email to deal with the funds...

http://www.intactaudio.com/contact.html
Get Your Fix
www.hifiheroin.com
Post Reply