Hello guys,
My current project is to build a speaker+amp set for my desktop.
I'm currently building a pair of speakers with the MarkAudio Alpair 6.2M with the planset from Planet10 Hifi.
The goal is to power it with a SE tube amp mainly inspired from Tubelab SSE. I already have a 12AT7 and a 6L6G in stock (just need to buy one more 😉).
I'd like to go with the Elcor GXSE15-5K output transformer. Since they are build to order, I want to order them ASAP. I already cut all the speakers panels !
What secondary ouput would you recommend ?
The Alpair has a ReVc of 3.6ohms but the impedance in AC looks like this :
Edcor let us choose between 4, 6, 8, 16ohms. I was first thinking that it would be 4ohms to match ReVc but looking at this graph, I see that it never hits 4ohms... It's more around 5ohms and more.
What do you think ?
Michaël
My current project is to build a speaker+amp set for my desktop.
I'm currently building a pair of speakers with the MarkAudio Alpair 6.2M with the planset from Planet10 Hifi.
The goal is to power it with a SE tube amp mainly inspired from Tubelab SSE. I already have a 12AT7 and a 6L6G in stock (just need to buy one more 😉).
I'd like to go with the Elcor GXSE15-5K output transformer. Since they are build to order, I want to order them ASAP. I already cut all the speakers panels !
What secondary ouput would you recommend ?
The Alpair has a ReVc of 3.6ohms but the impedance in AC looks like this :
Edcor let us choose between 4, 6, 8, 16ohms. I was first thinking that it would be 4ohms to match ReVc but looking at this graph, I see that it never hits 4ohms... It's more around 5ohms and more.
What do you think ?
Michaël
Oh OK.
Here is the link to the datasheet :
https://www.markaudio.com/wp-content/uploads/2019/07/Alp6M_Gen.2-june2011.pdf
I've made research and it seems people already talked about the Alpair 6.2M impedance matching better with a 6ohms secondary output impedance.
Here is the link to the datasheet :
https://www.markaudio.com/wp-content/uploads/2019/07/Alp6M_Gen.2-june2011.pdf
I've made research and it seems people already talked about the Alpair 6.2M impedance matching better with a 6ohms secondary output impedance.
Interesting. ReVc = 3.6ohms, but "Driver" = 4.0 ohms, and then (as you say) the impedance curve tells a different story.
I'd be inclined to go with 6ohm secondaries based on the curve and based on your report that others have recommended that as well. However, I am no expert on these matters.
I'd be inclined to go with 6ohm secondaries based on the curve and based on your report that others have recommended that as well. However, I am no expert on these matters.
Yes, at low frequencies. Use a shunt RC on the driver to flatten the HF impedance.
Either a 5 or 6 ohm secondary should work, I would get 5 ohms and use the Zobel.
Either a 5 or 6 ohm secondary should work, I would get 5 ohms and use the Zobel.
Last edited:
The data sheet shows graphs for the Gen 1 driver, but the the text on the front page talks about all the improvements (changes) in the Gen 2. So the Gen 2 driver will not match the curves, and the curves will indeed change when the driver is placed in a box.
The impedance bump at the driver's resonant frequency is normal. Depending on the box tuning that bump may move, get larger, or two distinct peaks may form. The impedance in the bass region will usually be larger than the "nominal" impedance spec except at very low frequencies, where you don't want to operate these speakers due to Xmax limitations.
This driver will be roughly 5 ohms from 300 Hz to 1500 Hz. That's where the "meat" of most music lies.
Connecting a 5 ohm load to the 6 ohm winding will reflect a 4167 ohm load to the output tube. A 6L6G will drive this load, but the THD will be a bit higher than normal, and the damping factor will be a bit worse.
Connecting a 5 ohm load to the 4 ohm winding will reflect a 6250 ohm to the output tube. This load will offer better THD and damping factor than the 4167 ohm load, but you will give up about 20% of the output power over the 4167 ohm load.
In the bass and treble regions the load on the output tubes are well above any limitations in the tube's ability to drive it.
If you plan to run only 6L6 type tubes in this amp and will have a B+ voltage over 400 volts, go with the 4 ohm version. The 6 ohm version will strain the tubes a bit in the frequency range where you will hear it most. If your B+ is less than 400 volts the 6 ohm version is better since the tubes will not be stressed as much, and you will be short on power output.
If you ever plan on using EL34's, KT88's or 6550's get the 6 ohm version. Those tubes can drive 4167 ohms with ease.
The impedance bump at the driver's resonant frequency is normal. Depending on the box tuning that bump may move, get larger, or two distinct peaks may form. The impedance in the bass region will usually be larger than the "nominal" impedance spec except at very low frequencies, where you don't want to operate these speakers due to Xmax limitations.
This driver will be roughly 5 ohms from 300 Hz to 1500 Hz. That's where the "meat" of most music lies.
Connecting a 5 ohm load to the 6 ohm winding will reflect a 4167 ohm load to the output tube. A 6L6G will drive this load, but the THD will be a bit higher than normal, and the damping factor will be a bit worse.
Connecting a 5 ohm load to the 4 ohm winding will reflect a 6250 ohm to the output tube. This load will offer better THD and damping factor than the 4167 ohm load, but you will give up about 20% of the output power over the 4167 ohm load.
In the bass and treble regions the load on the output tubes are well above any limitations in the tube's ability to drive it.
If you plan to run only 6L6 type tubes in this amp and will have a B+ voltage over 400 volts, go with the 4 ohm version. The 6 ohm version will strain the tubes a bit in the frequency range where you will hear it most. If your B+ is less than 400 volts the 6 ohm version is better since the tubes will not be stressed as much, and you will be short on power output.
If you ever plan on using EL34's, KT88's or 6550's get the 6 ohm version. Those tubes can drive 4167 ohms with ease.
Thank you all for your answers, especially George 😉.
I take the opportunity to thank you for your excellent website !
I too saw the mismatch in revision numbers, but this is the official link to the datasheet on the Alpair 6.2M page...
I must say, I don't know what to do now 😀 !
I already have one 6L6G (Tung Sol) in stock. That's why I was thinking about it.
But if building an amp around it means having less possibilities in term of tubes... well it makes me reconsider.
By the way, I assume that it would be best to have matched tubes right ?
I take the opportunity to thank you for your excellent website !
I too saw the mismatch in revision numbers, but this is the official link to the datasheet on the Alpair 6.2M page...
I must say, I don't know what to do now 😀 !
I already have one 6L6G (Tung Sol) in stock. That's why I was thinking about it.
But if building an amp around it means having less possibilities in term of tubes... well it makes me reconsider.
By the way, I assume that it would be best to have matched tubes right ?
Another question that I'm asking myself for a long time :
Since all speakers have a big bump in the impedance chart around their Fc, how do tube amp manage that ? The output transformer have always a fixed ratio (obviously) so does it means that they never handle the frequencies around Fc well ?
This Alpair is no really bad if you consider a 20/5 = 4 ratio between max and min impedance. A lot of 8ohms speakers do 80/6.5 = 12 ratio. Since this reflects on the primary of the OT, I wonder what's going on then.
This is not the same thing for the high end since we can correct the impedance with a RC network parallel to the speaker.
Since all speakers have a big bump in the impedance chart around their Fc, how do tube amp manage that ? The output transformer have always a fixed ratio (obviously) so does it means that they never handle the frequencies around Fc well ?
This Alpair is no really bad if you consider a 20/5 = 4 ratio between max and min impedance. A lot of 8ohms speakers do 80/6.5 = 12 ratio. Since this reflects on the primary of the OT, I wonder what's going on then.
This is not the same thing for the high end since we can correct the impedance with a RC network parallel to the speaker.
While a transformer can have a set output impedance like 4 or 8 the speaker impedance is important too. It is the reflected impedance back to primary that is important as well.
A 4 ohm output impedance and a 8 ohm speaker will raise primary impedance significantly.
The reflected impedance is turns ratio squared times speaker impedance.
But as usual I could be wrong ! but that's how I see it until corrected.
A 4 ohm output impedance and a 8 ohm speaker will raise primary impedance significantly.
The reflected impedance is turns ratio squared times speaker impedance.
But as usual I could be wrong ! but that's how I see it until corrected.
OTOH the speaker "handles" this range EXTRA well. It does not need so many Watts to do its work.
Read Radiotron Designer's Handbook, "The Relation Between the Output and the Loudspeaker".
This is when damping comes into play. Depending on the output impedance of the amplifier and the actual design of the speaker, you could end up with a serious problem in that area or no problems at all.
There are limits to how much "extra" impedance the OT can reflect. If we assume your 6L6 output tube is triode strapped, then you can crunch the numbers to calculate the extra output from changing from the stated impedance of your OT to five times that impedance, which is likely far beyond what the OT would do anyway. Remember that down around where that resonance is, the primary inductance of the output transformer will limit the actual load impedance.
For a full range driver that may roll off a little on top, the slight boost at high frequencies that you could get might not be a bad thing.
George, have to throw out a point or two. Since this is all SE, and running Class A, I am not quite going to buy into an EL34 being able to drive things easier than a 6L6. The idle current, and plate resistance of both are comparable. The load is set by the envelope delivered by the g2 voltage effect on the plate curves. Same sort of thing going on between a sweep tube and a 6V6...the g1=0V line of a sweep tube can be radically higher current than a 6V6, so a lower load( numeric ) to fit to these plate curves is required. One can indeed run a sweep tube amp into a large load, but the g2 voltage is going to be significantly lower than what you'd apply to a 6L6( assuming the delivered plate dissipation is the same ). A 31JS6C will want something well under 100V where a 6L6GC will want something closer to 300.
This example is slightly complicated by the fact that the JS6C has a lot lower plate resistance than a 6L6GC, and that the likely actual plate dissipation is higher. It remains that the low voltage g1=0V curve should intersect the load before the knee starts bunching stuff up appreciably though... 🙂
With resistive loads, running various pentodes as voltage amplifiers, the g2 effect on the envelope the load has to swing through is critical, and not so effected by frequency( the OPT's inductance, or as is often the case, lack there of ).
cheers,
Douglas
This example is slightly complicated by the fact that the JS6C has a lot lower plate resistance than a 6L6GC, and that the likely actual plate dissipation is higher. It remains that the low voltage g1=0V curve should intersect the load before the knee starts bunching stuff up appreciably though... 🙂
With resistive loads, running various pentodes as voltage amplifiers, the g2 effect on the envelope the load has to swing through is critical, and not so effected by frequency( the OPT's inductance, or as is often the case, lack there of ).
cheers,
Douglas
Member
Joined 2009
Paid Member
I think more important will be the quality of the transformer than a few ohms differences. I would suggest you get something around 8R because it will provide better damping and lower distortion than 6R or 4R and you’ll not notice the loss in power.
dont sweat it too much. You will be building new speakers before you know it. come to think about it the amp will prolly get replaced soon too. Its just the way this stuff works 🙂
Thanks all for your comments !
After a week I decided to :
- wait for the Alpair to arrive, then measure their impedance with LIMP
- I discovered that there's a company which build custom transformers a few km from me ! So I think I'll work with them. They can do custom power transformer, self and OT (with multiple secondaries)
- I'll probably go for a pair of matched tubes. It's not that expensive vs a single 6L6G plus I asked the company that sell the 6L6G to me and currently their tubes have a larger Ip and Gm that mine. Plus I saw that the KT88 will be able to supply more power with less distortion.
After a week I decided to :
- wait for the Alpair to arrive, then measure their impedance with LIMP
- I discovered that there's a company which build custom transformers a few km from me ! So I think I'll work with them. They can do custom power transformer, self and OT (with multiple secondaries)
- I'll probably go for a pair of matched tubes. It's not that expensive vs a single 6L6G plus I asked the company that sell the 6L6G to me and currently their tubes have a larger Ip and Gm that mine. Plus I saw that the KT88 will be able to supply more power with less distortion.