Hi,
The absolute best for MC headamp application is IMHO the E283CC a dual triode frame grid valve.
Not linear at Vg=0 though.
EC86 and EC88 are also fine candidates but are single triodes that can be used with a grounded cathode and no input coupling cap. Same goes for ECC88.
Cheers,
The absolute best for MC headamp application is IMHO the E283CC a dual triode frame grid valve.
Not linear at Vg=0 though.
EC86 and EC88 are also fine candidates but are single triodes that can be used with a grounded cathode and no input coupling cap. Same goes for ECC88.
Cheers,
Last edited:
Hi Everyone,
I recently updated my list of noise measurements to include the JJ 12AY7 and the new JJ 6SL7. I only measured 3 of each type, so these results are preliminary.
Although the 12AY7 is described in many tube manuals as "low noise and microphonics", the input-referred noise I measured (0.725 uV RMS) puts it in the middle of the list. One of the 3 samples had popcorn noise.
On average, the JJ 6SL7 noise performance seems ~ 1dB better than the Tung Sol Reissue 6SL7GT, but not quite as good as the NOS Sovtek 6H9C/6SL7GT. Still, I welcome this tube as a 2nd current production 6SL7 (the re-branded Sovteks are NOS), and much lower cost than the Tung Sol Reissue. I'll be measuring more.
Cheers,
Scott
I recently updated my list of noise measurements to include the JJ 12AY7 and the new JJ 6SL7. I only measured 3 of each type, so these results are preliminary.
Although the 12AY7 is described in many tube manuals as "low noise and microphonics", the input-referred noise I measured (0.725 uV RMS) puts it in the middle of the list. One of the 3 samples had popcorn noise.
On average, the JJ 6SL7 noise performance seems ~ 1dB better than the Tung Sol Reissue 6SL7GT, but not quite as good as the NOS Sovtek 6H9C/6SL7GT. Still, I welcome this tube as a 2nd current production 6SL7 (the re-branded Sovteks are NOS), and much lower cost than the Tung Sol Reissue. I'll be measuring more.
Cheers,
Scott
Tavishdad, how about these tubes (used in mic circuits)
- the 'low noise' pentode sylvania 7543 (triode mode)
- 'vibration resistant' pentode ge5654
The ge5654 in particular is widely available.
ps - thanks for your work
I used the jj5751 recently with much measuring and indeed found it to be the lowest in noise of all my similar tubes.
- the 'low noise' pentode sylvania 7543 (triode mode)
- 'vibration resistant' pentode ge5654
The ge5654 in particular is widely available.
ps - thanks for your work
I used the jj5751 recently with much measuring and indeed found it to be the lowest in noise of all my similar tubes.
Thanks. I've been intending to measure the 6AU6 & 6AK5, so including a few of those variants you mention will be straightforward. I'll probably start with the 6AU6, since I have quite a few on hand, and I'm curious how it compares to the EF86. Scott
Tavishdad, how about these tubes (used in mic circuits)
- the 'low noise' pentode sylvania 7543 (triode mode)
- 'vibration resistant' pentode ge5654
The ge5654 in particular is widely available.
ps - thanks for your work
I used the jj5751 recently with much measuring and indeed found it to be the lowest in noise of all my similar tubes.
Hmm. I looked at a spec sheet. Interesting. It's not obvious to me which external electrodes are the plate, grid, and filament, however.
These seem to be operated with zero bias. Have you ever measured the input impedance, or grid current?
Thanks, Scott
Any results for subminiature tubes/valves?
Hi,
I saw a post mentioning Russian 6N16B and 6C6B (if I recall correctly), do you have plans to test, for example:
6C7B triode
6N17B triode
6N21B triode
5744 triode
These 're all available on fleaBay for relatively cheaply.
I am particularly interested in noise measurements of the 5744 /WB triode which I'm using for several projects in the works. (It's touted by vendors as being a subminiature 12AX7, how correct that is I don't know. Mu of 70 and anode current are similar)
I may be able to post a tube for you to test, if postage is reasonable.
Hi,
I saw a post mentioning Russian 6N16B and 6C6B (if I recall correctly), do you have plans to test, for example:
6C7B triode
6N17B triode
6N21B triode
5744 triode
These 're all available on fleaBay for relatively cheaply.
I am particularly interested in noise measurements of the 5744 /WB triode which I'm using for several projects in the works. (It's touted by vendors as being a subminiature 12AX7, how correct that is I don't know. Mu of 70 and anode current are similar)
I may be able to post a tube for you to test, if postage is reasonable.
Last edited:
Suppose we want a low noise tube for the input stage of a Moving Coil Cartridge. What factors might we consider?
A Moving Coil Cartridge might have 0.1mV (0.0001V) output for 5cm/sec velocity. It might have only 10 Ohms DC resistance. The Preamp might have a 90 Ohm input resistance setting.
0.0001V / 100 Ohms = 1uA current in the 10 Ohm coil and the 90 Ohm input resistor. (The loss using a 90 Ohm resistor versus using an open circuit is 1 dB).
Now, the record has a loud section, 25cm/sec velocity. The cartridge output is 5uV, and the current is 5uA.
Hopefully, the input tube we selected will be less than 100 nano Amp grid current with no signal in. (0.1 uA).
What is the problem with Zero Bias, if the maximum signal is 5uV from 9 Ohms (10 Ohms in parallel with 90 Ohms)?
If the input tube has u = 100, and a current source for the plate load, and the next stage grid resistor is extremely high, the gain will approach 100.
5uV x 100 = 500uV (0.5mV). The plate will swing +/- 0.5mV on that large 25cm/sec signal. In my opinion, the “linearity” of the input tube is not an issue in this case.
A Moving Coil Cartridge might have 0.1mV (0.0001V) output for 5cm/sec velocity. It might have only 10 Ohms DC resistance. The Preamp might have a 90 Ohm input resistance setting.
0.0001V / 100 Ohms = 1uA current in the 10 Ohm coil and the 90 Ohm input resistor. (The loss using a 90 Ohm resistor versus using an open circuit is 1 dB).
Now, the record has a loud section, 25cm/sec velocity. The cartridge output is 5uV, and the current is 5uA.
Hopefully, the input tube we selected will be less than 100 nano Amp grid current with no signal in. (0.1 uA).
What is the problem with Zero Bias, if the maximum signal is 5uV from 9 Ohms (10 Ohms in parallel with 90 Ohms)?
If the input tube has u = 100, and a current source for the plate load, and the next stage grid resistor is extremely high, the gain will approach 100.
5uV x 100 = 500uV (0.5mV). The plate will swing +/- 0.5mV on that large 25cm/sec signal. In my opinion, the “linearity” of the input tube is not an issue in this case.
I was surprised to see this old thread pop up again. It's been a while since I've added any new tubes to the list, although my son and I screened a LOT of JJ 5751 and Tung Sol Reissue 6SL7GT over the past few years.
I may measure more tubes in the future, so I'll put those on my list to consider. The 5744 looks interesting - my tube manuals list it as a Raytheon UHF triode. But with no other information, I wouldn't expect it to have particularly low noise in the audio range. The only subminiature tubes I've measured (the RU-6N16 and the triode-connected 5840) were above 11dB NF in my test setup. And with few exceptions, NOS tubes tend to be noisy.
One thing I've learned is that even quiet tubes like the JJ 5751 occasionally come in noisy batches, so tube quality control these days isn’t the best. It’s necessary to measure a lot of tubes from several batches to be really sure about a brand and type. The Tung Sol Reissue 6SL7GT is particularly prone to noisy samples and batches. It’s still a good tube, I just wish their quality control was better.
Scott
I may measure more tubes in the future, so I'll put those on my list to consider. The 5744 looks interesting - my tube manuals list it as a Raytheon UHF triode. But with no other information, I wouldn't expect it to have particularly low noise in the audio range. The only subminiature tubes I've measured (the RU-6N16 and the triode-connected 5840) were above 11dB NF in my test setup. And with few exceptions, NOS tubes tend to be noisy.
One thing I've learned is that even quiet tubes like the JJ 5751 occasionally come in noisy batches, so tube quality control these days isn’t the best. It’s necessary to measure a lot of tubes from several batches to be really sure about a brand and type. The Tung Sol Reissue 6SL7GT is particularly prone to noisy samples and batches. It’s still a good tube, I just wish their quality control was better.
Scott
Hi TavishDad,
Well I saw thus thread pop up recently and it piqued my interest and I read a fair way through it, but I honestly didn't notice it's age!
The 5744 submini is described as a UHF triode with is also useful for AF voltage amplifier duty. Though I guess that doesn't describe the tube as low noise (at least not it's primary attribute)
Thanks for you reply anyway
Well I saw thus thread pop up recently and it piqued my interest and I read a fair way through it, but I honestly didn't notice it's age!
The 5744 submini is described as a UHF triode with is also useful for AF voltage amplifier duty. Though I guess that doesn't describe the tube as low noise (at least not it's primary attribute)
Thanks for you reply anyway
Hi there... I have several 6S17K-V and ECC88 tubes, a uTracer, a balanced 24Bit 192Khz sound card, a dual channel 100MHz digital Tek scope, a Salas High Voltage Shunt Regulator and a free weekend.
My plan is to use a combination of 6C17K-V and ECC88 to build a 60dB minimum head amp for LOMC cartridge. For now I'm thinking of a differential cascode or two LTPs dc coupled with Norton level shifters. It works in LTSpice
Any hints on how to make the LTSpice simulations work in real life?
Silviu
Last edited:
The link on the first page of this thread still takes you to the most recent version of my list, which includes a lot of details of the setup. There is a pdf file you can download, or you can scroll through the list and explanation online. Scott
I don't remember measurement results, but it biases itself pretty well by grid current through 47k grid leak resistor. 10k in anode, 200V B+
However, they are hard to match, so I use them soldered on octal tube plugs, with microphone grills on top.
I built the measurement amplifier "dead bug" style but with OPA2604 which are somewhat lower noise than the LF353. This is pretty much what one would expect from the datasheet, with A=100. The left axis in V/RtHz. slightly different in that the gain is split into two sections. Measuring the RMS voltage with the 30-8000Hz filter we get 108uV, or about 12.2nV/Rt Hz
Attachments
Last edited:
Hi Jack,
I'd be delighted to have someone else making these tube noise measurements, too! I could probably send you a pcb for the noise measurement amplifier - I'd just have to dig one out from under the piles, pretty sure I still have a blank one.
I could also measure a few tubes and send them to you. You could remeasure them and we'll see if we get the same answer. I have a carton of JJ 5751 "rejects" from sorting, too noisy to use in a phono preamp, but ideal for a cross check of our setups.
Scott
I'd be delighted to have someone else making these tube noise measurements, too! I could probably send you a pcb for the noise measurement amplifier - I'd just have to dig one out from under the piles, pretty sure I still have a blank one.
I could also measure a few tubes and send them to you. You could remeasure them and we'll see if we get the same answer. I have a carton of JJ 5751 "rejects" from sorting, too noisy to use in a phono preamp, but ideal for a cross check of our setups.
Scott
Scott:
I have oddball tubes -- like 20EZ7's and others I wanted to subject to the same analysis.
Will not have a lot of time with the soldering iron until late March but will get back to you.
Jack
You may be over-estimating the noise since you're using a 20dB/8vo low pass filter. On the other hand, is the 220k resistor R7 purpose to discharge C4?
I have oddball tubes -- like 20EZ7's and others I wanted to subject to the same analysis.
Will not have a lot of time with the soldering iron until late March but will get back to you.
Jack
You may be over-estimating the noise since you're using a 20dB/8vo low pass filter. On the other hand, is the 220k resistor R7 purpose to discharge C4?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.