Dr Curl how youre doin
if i may i want to ask you a couple of questions
what's the s/n of the input stage of the blowtorch (in isolation to
the other stages) and what's the transconductance of the input
devices and what percentage of matching have you achieved with
the input devices, are they special grade
thanks
if i may i want to ask you a couple of questions
what's the s/n of the input stage of the blowtorch (in isolation to
the other stages) and what's the transconductance of the input
devices and what percentage of matching have you achieved with
the input devices, are they special grade
thanks
Folks, having big problems with my computer, so I may be intermittent in response. The 2N4401-3 was first made by Motorola. Now, many manufacturers make it. I don't design ground planes, so I can't comment.
Matching is done by Toshiba, and servos make up for any offset. This is not the lowest noise possible, because of an added contribution from the mosfets in the second stage. It is ok to follow a 10-25K volume control.
Matching is done by Toshiba, and servos make up for any offset. This is not the lowest noise possible, because of an added contribution from the mosfets in the second stage. It is ok to follow a 10-25K volume control.
The 2N4401-4403 series were mentioned in Motchenbacher and Fitchen's (sp?) book "Low Noise Electronics" back in the 70's, a great book on the charateristics of noise and low noise electronic design. Maybe that's where the transistors got their reputation. Marshall Leach also mentions them (and uses them) in his moving coil preamp design.
My sympathies, Mr. Curl. My home computer took a big dump last Sunday, and I've been scurrying aroung trying to put things back together. I finally took the plunge and bought both a UPS and an external hard drive for backup.
My sympathies, Mr. Curl. My home computer took a big dump last Sunday, and I've been scurrying aroung trying to put things back together. I finally took the plunge and bought both a UPS and an external hard drive for backup.
Well tried , JustCallMeDad...
LOL ! But no class when the teacher is out, I am afraid !
Jbaudiophile
LOL ! But no class when the teacher is out, I am afraid !
Jbaudiophile
I first discovered the low noise properties of the 2n4401-3 transistors when working on low noise electronics at Ampex in 1968, soon after these devices were released by Motorola. I had a Quantech noise tester available, and I tested everything that was released that looked promising. It wasn't a well known low noise part, until M&F published their great book on low noise design in 1973. It was used in the Levinson JC-1 pre-preamp, that was first released in 1973. For years this was the best part-pair, but in the late 70's Hitachi made a really super part-pair, that is still the best ever done, as far as I can tell. However they discontinued the parts, decades ago.
The best 'modern' paper on semiconductor noise is: 'The Design of Low-Noise Amplifiers' By Yishay Netzer, 'Proceedings of the IEEE' Vol. 69, NO 6, June 1981.
This is a good example of: 1. It doesn't exist. 2. It exists, but it is not important. 3. We invented it, or in this case, it is obvious. 🙂
The best 'modern' paper on semiconductor noise is: 'The Design of Low-Noise Amplifiers' By Yishay Netzer, 'Proceedings of the IEEE' Vol. 69, NO 6, June 1981.
This is a good example of: 1. It doesn't exist. 2. It exists, but it is not important. 3. We invented it, or in this case, it is obvious. 🙂
The intersting question would be who makes the best 4401s and 4403s, as parameters like popcorn noise depend on processing cleanliness. Being a cheapskate, I use whatever I can get from local surplus houses or Ebay. Living in Silicon Valley, I see NOS devices from National, Fairchild (the old Fairchild), Motorola, Rohm, Synertek (?!!) (no TI). I think GE and Sprague also made 4401s and 4403s for a while. If I had a noise analyzer, it would be interesting to root through my old stock and look at the variation between vendors. I can't say whose new devices would be best, though I would probably steer clear of the new Farchild devices on general principle. New devices from any vendor may be a crap shoot, as they may actually be made in some sleazy outside fab and just marked with a major logo. Profit margin in small signal stuff is very small these days. This is no doubt why companies like National and TI dumped their small signal lines.
There are MANY better parts than the 4401-3 pair. It is now 38 years old! Everyone, please use better parts.
Help
John,
You mentioned some Hitachi parts that might be the best ever. Which pair is this?
I find time to search through old surplus shops. It is surprising what was purchased by NASA, Honeywell, and other high tech companies 20 -30 years ago. It sometimes ends up in these places for pennies on the dollar.
George
John,
You mentioned some Hitachi parts that might be the best ever. Which pair is this?
I find time to search through old surplus shops. It is surprising what was purchased by NASA, Honeywell, and other high tech companies 20 -30 years ago. It sometimes ends up in these places for pennies on the dollar.
George
Hey George
Are there any surplus shops other than Ace that I don't know about in the NASA area?
Are there any surplus shops other than Ace that I don't know about in the NASA area?
Shops
Brian,
EPO is on Fondren, there is another used to be EPO at Hiway 3 and Medical Center in Webster, Tanner off 35 in Dallas, and BGMcro in Dallas also.
The stock is getting low, but EPO on Fondren had Vishay S102 for 20 cents each. Picked up some Vishay 1280G trimmers yesterday for 1.50 each at another shop.
Nedless to say I am building all my stuff with the S102 right now.
George
Brian,
EPO is on Fondren, there is another used to be EPO at Hiway 3 and Medical Center in Webster, Tanner off 35 in Dallas, and BGMcro in Dallas also.
The stock is getting low, but EPO on Fondren had Vishay S102 for 20 cents each. Picked up some Vishay 1280G trimmers yesterday for 1.50 each at another shop.
Nedless to say I am building all my stuff with the S102 right now.
George
I knew about EPO on Fondren, but never found goodies there. It's mostly robotics, overpriced comuter crp and boom box. Very few fets and BPtransistors. I'll have to try the one in Webster. Not much can match up to Ace.
To get things on track, I might point out a few things about low noise transistors.
First, transistors have 3 significant sources that must be minimized, if possible. These are: voltage noise, current noise, and base resistance.
Many 'low noise' transistors have high beta. This is important to reduce current noise, which is proportional to base current. However, high beta usually also means high base resistivity, so these devices are always more noisy with low source resistances, such as moving coil cartridges.
The 4401-3 is a low-medium beta device, so it is almost worthless for line amps, power amps, etc, BUT it has one major advantage: It has low base resistance of about 40 ohms, rather than 400 ohms, typical of early low noise devices.
Today, there are many devices, usually Japanese, that are pretty good general purpose parts. I don't have specific numbers, because I almost always use fets, as they are just as quiet as a 4401-3, at every input impedance.
Please remember, the 4401-3 was just about all there was for low inpedance design, for about 10 years, maybe from 1968-1978. Then Japan made better parts, Hitachi made a super part, but they never got released openly, so it would probably be a waste of time for any of you to try to find them. Rohm has made many good low noise parts, but some of these are hard to get, as well. For serious low noise design, it is best to use fets, instead, or maybe a Rohm part, if you can find them.
First, transistors have 3 significant sources that must be minimized, if possible. These are: voltage noise, current noise, and base resistance.
Many 'low noise' transistors have high beta. This is important to reduce current noise, which is proportional to base current. However, high beta usually also means high base resistivity, so these devices are always more noisy with low source resistances, such as moving coil cartridges.
The 4401-3 is a low-medium beta device, so it is almost worthless for line amps, power amps, etc, BUT it has one major advantage: It has low base resistance of about 40 ohms, rather than 400 ohms, typical of early low noise devices.
Today, there are many devices, usually Japanese, that are pretty good general purpose parts. I don't have specific numbers, because I almost always use fets, as they are just as quiet as a 4401-3, at every input impedance.
Please remember, the 4401-3 was just about all there was for low inpedance design, for about 10 years, maybe from 1968-1978. Then Japan made better parts, Hitachi made a super part, but they never got released openly, so it would probably be a waste of time for any of you to try to find them. Rohm has made many good low noise parts, but some of these are hard to get, as well. For serious low noise design, it is best to use fets, instead, or maybe a Rohm part, if you can find them.
- Status
- Not open for further replies.