Something weird happened with my previous post and it can't be edited, deleted, or replied to, so I'm trying again.
Wondering what specs for Electra-Print OPTs you'd recommend for a TSE 300b build, given that my speakers are as follows:
JM Lab Electra 905
My instinct is to go for higher B+ and input impedance for better distortion and damping, but a lot of builders seem to have high efficiency horns so I'm not sure if my setup would require a more optimal configuration.
Jack has radically different suggestions than what people tend to do for a TSE, so I'd like to poll experienced builders.
I'm eager to get this settled so I can start the project!
Thanks,
Jeff
Wondering what specs for Electra-Print OPTs you'd recommend for a TSE 300b build, given that my speakers are as follows:
JM Lab Electra 905
- 2-way MTM
- 91.5 db Efficiency
- 8 ohm impedance
My instinct is to go for higher B+ and input impedance for better distortion and damping, but a lot of builders seem to have high efficiency horns so I'm not sure if my setup would require a more optimal configuration.
Jack has radically different suggestions than what people tend to do for a TSE, so I'd like to poll experienced builders.
I'm eager to get this settled so I can start the project!
Thanks,
Jeff
Jeff,
Jack is no fan of the TSE design. You will find this thread useful (and amusing):
http://www.diyaudio.com/forums/tubelab/272474-damm-not-response-i-expected.html
Btw, George does state that his personal preference is to use 5k OPTs and run 300B tube B+ closer to 400v:
Tubes and Applications | Tubelab
He mentions the bias for 300Bs between 60-75 ma. Some builders run certain tubes (like the JJ 300B) at higher bias points too (80-90 ma).
Tough question on your speakers - 91.5 dB on paper looks workable with a 7-8 watt 300B amp, however no idea if the speakers are difficult to drive and need a more powerful amplifier. If you had higher sensitivity speakers, there would be more headroom with the same amp.
Jack is no fan of the TSE design. You will find this thread useful (and amusing):
http://www.diyaudio.com/forums/tubelab/272474-damm-not-response-i-expected.html
Btw, George does state that his personal preference is to use 5k OPTs and run 300B tube B+ closer to 400v:
Tubes and Applications | Tubelab
He mentions the bias for 300Bs between 60-75 ma. Some builders run certain tubes (like the JJ 300B) at higher bias points too (80-90 ma).
Tough question on your speakers - 91.5 dB on paper looks workable with a 7-8 watt 300B amp, however no idea if the speakers are difficult to drive and need a more powerful amplifier. If you had higher sensitivity speakers, there would be more headroom with the same amp.
Last edited:
Thanks zman01. These speakers are easy to drive. Years ago I heard them on a pair of 7/8 watt SE 300b mono blocks and it changed my life. So I'm reasonably confident the TSE will drive them well. I was primarily wondering if people with speakers like mine (on paper) had bad results with the 5k OPTs and their lower power, and if maybe I should go for a 3k or 3.5k OPT.
The TSE was designed over 10 years ago. At that time there were NO reasonably priced 300B tubes that reliably worked in the 400 volt range. Sovtek claimed that their tubes worked to 400 and even 450 volts, and SOME did, but others would go into runaway. This was because of a poorly aligned grid structure.
The TSE was designed to operate at 360 volts and below and is a rather compact design. Many builders have used the TSE in the 400 to 500 volt range, and the circuit design will work at those voltages, but good ventilation is needed around the mosfets at those voltages.
I still use my original TSE. It still has the original PC board that I cooked up on the kitchen stove (not the shiny new ones from the PC board vendor) and the Electra Print transformers. It is Tubelab amp #1 (the first one I built) and still the best sounding low power amp I have. The transformers are 5K ohm to 8 ohm 10 watt EP's. Just don't call them Albert's!
I looked for an impedance VS frequency graph on your speakers and couldn't find one via a Google search. If you have, or can get this graph it will help with your decision.
The transformer has an impedance spec, (like 5K to 8 ohms) but it transforms a range of impedances using the same ratio. In this manner an 8 ohm speaker becomes a 5K load to the 300B, but a 7 ohm speaker becomes a 4375 ohm load to the tube, and a 9 ohm speaker becomes a 5625 ohm load.
An 8 ohm speaker is NOT an 8 ohm speaker at all frequencies. In fact it may not be an 8 ohm speaker at all. The 8 ohm rating is "nominal". Many manufacturers interpret this to mean 8 ohms at 1 KHz. Some may assign a different frequency, and some choose not to disclose ANY impedance information at all. My Yamaha NS-10M Studio monitors came with an impedance curve. They are about 8 ohms at 2 KHz and above 8 ohms at most other frequencies. They are over 20 ohms below 100 Hz, as are most speakers with small woofers.
If your speakers have a significant portion of the impedance curve below 8 ohms at frequencies where there is significant musical energy present in your preferential listening choices, then I would tend toward the 5K ohm transformers. If most of the curve is above 8 ohms in the range of interest, them 3.6K may be a better choice.
You say that you heard a 300B amp through your speakers that sounded good. Can you find out what transformers it used? The transformers interact with the speakers to create the biggest influence on how the overall system sounds. The output tube choice also makes a significant contribution.
The load that a SE triode sees will determine its distortion, dynamics, and power output capabilities. Within a reasonable range a lower impedance load will increase the power and distortion, and decrease the accuracy and dynamic capability.
A 5K load on a 300B will typically give up 1 or 2 watts compared to a 3.6K load for a given B+ voltage. It will also improve dynamics, damping factor, and distortion by a similar ratio (about 20to 30%). Even with my 87db Yamahas, I will trade those 2 watts for a slightly better sound.
The TSE was designed to operate at 360 volts and below and is a rather compact design. Many builders have used the TSE in the 400 to 500 volt range, and the circuit design will work at those voltages, but good ventilation is needed around the mosfets at those voltages.
I still use my original TSE. It still has the original PC board that I cooked up on the kitchen stove (not the shiny new ones from the PC board vendor) and the Electra Print transformers. It is Tubelab amp #1 (the first one I built) and still the best sounding low power amp I have. The transformers are 5K ohm to 8 ohm 10 watt EP's. Just don't call them Albert's!
I looked for an impedance VS frequency graph on your speakers and couldn't find one via a Google search. If you have, or can get this graph it will help with your decision.
The transformer has an impedance spec, (like 5K to 8 ohms) but it transforms a range of impedances using the same ratio. In this manner an 8 ohm speaker becomes a 5K load to the 300B, but a 7 ohm speaker becomes a 4375 ohm load to the tube, and a 9 ohm speaker becomes a 5625 ohm load.
An 8 ohm speaker is NOT an 8 ohm speaker at all frequencies. In fact it may not be an 8 ohm speaker at all. The 8 ohm rating is "nominal". Many manufacturers interpret this to mean 8 ohms at 1 KHz. Some may assign a different frequency, and some choose not to disclose ANY impedance information at all. My Yamaha NS-10M Studio monitors came with an impedance curve. They are about 8 ohms at 2 KHz and above 8 ohms at most other frequencies. They are over 20 ohms below 100 Hz, as are most speakers with small woofers.
If your speakers have a significant portion of the impedance curve below 8 ohms at frequencies where there is significant musical energy present in your preferential listening choices, then I would tend toward the 5K ohm transformers. If most of the curve is above 8 ohms in the range of interest, them 3.6K may be a better choice.
You say that you heard a 300B amp through your speakers that sounded good. Can you find out what transformers it used? The transformers interact with the speakers to create the biggest influence on how the overall system sounds. The output tube choice also makes a significant contribution.
The load that a SE triode sees will determine its distortion, dynamics, and power output capabilities. Within a reasonable range a lower impedance load will increase the power and distortion, and decrease the accuracy and dynamic capability.
A 5K load on a 300B will typically give up 1 or 2 watts compared to a 3.6K load for a given B+ voltage. It will also improve dynamics, damping factor, and distortion by a similar ratio (about 20to 30%). Even with my 87db Yamahas, I will trade those 2 watts for a slightly better sound.
This is fantastic, and will provide the info I need to make an informed decision. I love learning. Thanks, George!
While searching for the Electras earlier this year, I came across this German site that posted measurements. Looks like it doesn't include impedence vs frequency, but perhaps this will help? I've emailed the manufacturer for information.
JMlab Electra 905
I was unable to get Google translate to access the page, so here are the respective graphs and the parameters they display.
Amplitude response freestanding on stand
Amplitude response and harmonic distortion at 88 dB
Waterfall Chart
Electrical impedance and phase response
It's not clear to me what the two lines in the last graph represent, but the left and bottom axes show impedance and frequency. If the lower line is relevant, the bulk of it seems to be above 8ohm, in which case a 3.6K ohm OPT might be optimal?
But...
To answer your question about the 300Bs I heard before, they were a one-off: a pair of George Wright mono blocks originally designed for 2A3s that he modified to accept a 300B. The OPT was Magnequest DS-025 (2.5K: 4, 8, 16; can be air gapped up to 70mA -- see Magnequest Products - MQ Iron Listing/Pricing). I believe these one-offs later went into production as the Mono 7, a 300B designed to sound more like a 2A3 which put out 7 watts and ran a lower voltage than a traditional 300b.
When I heard these about 15 years ago, my friend and I listened to a late 1950s recording of Beethoven’s 9th on vinyl. You could pinpoint every instrument in the orchestra in 3D space, and after 15 minutes of overemotional speechlessness the first comment was something like “so that’s what it’s like to be German.”
Words are inadequate to convey what happened when we put on Black Sabbath, as it was more akin to eating raw steak and the smell of scorched earth than an ear-borne experience.
They got plenty loud and plenty real.
So, if the TSE with a 5K OPT would put out ~7 watts and would sound more refined than a 3.5K, I'm thinking I'd be happier with the 5K. But of course it's a very different circuit than what I heard before, so I'd still really appreciate some thoughts.
Really looking forward to building this project.
While searching for the Electras earlier this year, I came across this German site that posted measurements. Looks like it doesn't include impedence vs frequency, but perhaps this will help? I've emailed the manufacturer for information.
JMlab Electra 905
I was unable to get Google translate to access the page, so here are the respective graphs and the parameters they display.
Amplitude response freestanding on stand
Amplitude response and harmonic distortion at 88 dB
Waterfall Chart
Electrical impedance and phase response
It's not clear to me what the two lines in the last graph represent, but the left and bottom axes show impedance and frequency. If the lower line is relevant, the bulk of it seems to be above 8ohm, in which case a 3.6K ohm OPT might be optimal?
But...
To answer your question about the 300Bs I heard before, they were a one-off: a pair of George Wright mono blocks originally designed for 2A3s that he modified to accept a 300B. The OPT was Magnequest DS-025 (2.5K: 4, 8, 16; can be air gapped up to 70mA -- see Magnequest Products - MQ Iron Listing/Pricing). I believe these one-offs later went into production as the Mono 7, a 300B designed to sound more like a 2A3 which put out 7 watts and ran a lower voltage than a traditional 300b.
When I heard these about 15 years ago, my friend and I listened to a late 1950s recording of Beethoven’s 9th on vinyl. You could pinpoint every instrument in the orchestra in 3D space, and after 15 minutes of overemotional speechlessness the first comment was something like “so that’s what it’s like to be German.”
Words are inadequate to convey what happened when we put on Black Sabbath, as it was more akin to eating raw steak and the smell of scorched earth than an ear-borne experience.
They got plenty loud and plenty real.
So, if the TSE with a 5K OPT would put out ~7 watts and would sound more refined than a 3.5K, I'm thinking I'd be happier with the 5K. But of course it's a very different circuit than what I heard before, so I'd still really appreciate some thoughts.
Really looking forward to building this project.
Last edited:
I run Electra-Print transformers in my Damn Good 300B amp.
Specs: 15 W, 100 mA, 5 kΩ primary, 4 and 8 Ω secondary taps, copper for use with 300B.
Jack at Electra-Print can be rather crusty to deal with, but if you specify the transformer exactly as I just did above, you'll have no trouble getting what you need.
I run JJ 300B tubes at 400 V, 85 mA. Sounds fantastic and gives you about 10 W output power (3 % THD).
Higher voltage, higher primary impedance gives you lower THD and, in theory, slightly lower output power. To me that's a worthwhile tradeoff and I've never found the difference in power to be anything to concern myself with anyway.
Tom
An externally hosted image should be here but it was not working when we last tested it.
Specs: 15 W, 100 mA, 5 kΩ primary, 4 and 8 Ω secondary taps, copper for use with 300B.
Jack at Electra-Print can be rather crusty to deal with, but if you specify the transformer exactly as I just did above, you'll have no trouble getting what you need.
I run JJ 300B tubes at 400 V, 85 mA. Sounds fantastic and gives you about 10 W output power (3 % THD).
Higher voltage, higher primary impedance gives you lower THD and, in theory, slightly lower output power. To me that's a worthwhile tradeoff and I've never found the difference in power to be anything to concern myself with anyway.
Tom
Yep. That's where I landed as well - for the same reasons. My speakers are 85-87 dB efficient as well. 10 W is plenty of power for my needs.
Tom
The upper line is relative phase VS frequency with zero arbitrarily assigned to 2 KHz.
The lower line is impedance VS frequency. Your speakers are below 8 ohms from 120 Hz to about 700 Hz. These are the primary fundamental frequencies for many common musical instruments, as well as a large chunk of a male vocalists range. There will be considerable energy in this range, so I would go with the 5K transformers.
he double resonant peaks below 100 Hz imply that the two woofers are not identical, or there are some tricks in the crossover to extend the low frequency range.
I'm placing the order for resistors, capacitors, diodes etc on Digikey. Some of the Panasonic metal film resistors were not available but there are plenty of Stackpole ones with the same values and the same format (ie through hole) that are recommended by the system. These should be OK right?
I just purchased a TSE board and was looking around here for the PCB connectors you used. The double impedance peaks are typical of a ported box design.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.