Yes, the curves bottom out on the right side.
At least at lower impedances.
I was talking about ultra low Rbb transistors. They really need some current to show improvements over higher Rbb ones. I found many MC heads where idle was just too low.
It makes of cause so sense to run the input stage on high current when you have a cartridge like the DL103 with 40 Ohm DC impedance.
At least at lower impedances.
I was talking about ultra low Rbb transistors. They really need some current to show improvements over higher Rbb ones. I found many MC heads where idle was just too low.
It makes of cause so sense to run the input stage on high current when you have a cartridge like the DL103 with 40 Ohm DC impedance.
The question is : what is high bias current ?
Some think that 1mA is already high. Some examples are the Lindsley-Hood designs and early Self.
In the latest version of his preamp ( Elektor ) he has ramped it up a bit for a total of 4mA though. I think that is still too low when you have one of the very low impedances cartridges that come out now, say from Lyra ( Atlas, 4 Ohm ) or My Sonic Lab ( 2 Ohm and less ). Modern neodym magnets make that possible. The Atlas for example gives out nearly 1mV at 5cm/sec. That is of cause a fantastic combination for high S/N.
It would be a pity when the MC head destroys this.
Some think that 1mA is already high. Some examples are the Lindsley-Hood designs and early Self.
In the latest version of his preamp ( Elektor ) he has ramped it up a bit for a total of 4mA though. I think that is still too low when you have one of the very low impedances cartridges that come out now, say from Lyra ( Atlas, 4 Ohm ) or My Sonic Lab ( 2 Ohm and less ). Modern neodym magnets make that possible. The Atlas for example gives out nearly 1mV at 5cm/sec. That is of cause a fantastic combination for high S/N.
It would be a pity when the MC head destroys this.
Here are Syn08´s noise calculations. When you go over it you find enormous bias when Rbb is ultra low. He runs his 0.3nV/qHz bipolar head on 28mA.
I battled
with him for month but in the end i gave in.
Of cause distortion has to be considered too, especially in an open loop design.
Taking this into account i found the optimum at 5 to 10mA per transistor.
That works only when the PNP has a high early voltage.
For a transistor like the BC327 with Rbb of ca.30 Ohm it comes out at 2.5mA.
More then that and the early effect comes into play quite drastic.
I battled
with him for month but in the end i gave in.
Of cause distortion has to be considered too, especially in an open loop design.
Taking this into account i found the optimum at 5 to 10mA per transistor.
That works only when the PNP has a high early voltage.
For a transistor like the BC327 with Rbb of ca.30 Ohm it comes out at 2.5mA.
More then that and the early effect comes into play quite drastic.
Here's the components of noise contribution for a 70R system -- something like the Denon DL103 (40 ohms) and transistors with rbb = 30R
An externally hosted image should be here but it was not working when we last tested it.
I looked in my book, Low Noise Electronic Design, and found the formula for transistor voltage noise.
En=4ktRbb+2qIcRe**2
Re**2 means Re squared. The first term is the thermal noise in the base spreading resistance and the second term is the noise in the Shockley emitter resistance, given below.
Re=kt/(qIc)=1/Gm
You can't change the first term but the second term can be reduced by increasing the collector current, Ic. You can decrease the second term to much less than the first term so base spreading resistance noise dominates.
4ktRbb>>2qIcRe**2
If I did the algebra correctly you obtain this.
Ic>>0.026/(2Rbb) or approximately Ic>>1/(80Rbb)
En=4ktRbb+2qIcRe**2
Re**2 means Re squared. The first term is the thermal noise in the base spreading resistance and the second term is the noise in the Shockley emitter resistance, given below.
Re=kt/(qIc)=1/Gm
You can't change the first term but the second term can be reduced by increasing the collector current, Ic. You can decrease the second term to much less than the first term so base spreading resistance noise dominates.
4ktRbb>>2qIcRe**2
If I did the algebra correctly you obtain this.
Ic>>0.026/(2Rbb) or approximately Ic>>1/(80Rbb)
The DL103R is 14 Ohms. What's the point anyway? It's easy to get the background noise well below the noise when dropped into an unmodulated grove.
Hello --
I didnt even know this subject existed until doing search on the Web about measuring Rbb and saw this .....
i owned a Denon DL103 so that is what i used. Medium output voltage.
I just had access
)) to those transistors I used - so i used them. No attempt to make the worlds lowest noise MC step-up but it was very quiet in the home music system. I had a Denon step-up transformer to compare against for noise and distortion and response.
The important point I wanted to find out was -- if a lower than 100 Ohm cart load (the typical MC value used) made the sound have lower distortion as was reported via listening by Peter Moncrief, then could I make the input Z of a MC pre-preamp with a very low z and would it sound even better -- or at least better than what I had been using. What would it sound like to My ears. The passive Rload on the cart had been to make it sound 'cleaner' and lower distortion but, of course the cart output level was reduced as well. So it was just an interesting affect but not practicle.
It was developed to provide a very low input Z, direct coupled as much as possible and a sysmetrical design with the benefit of lower distortion coming off the cartrdige.
--- now that i dont even have a phono/LP any more --- to maximize its potential with better transistors etal. I gave it to Kavi, also. If I had a phono system I would love to hear what improvements you will make to it.
Eventually - try to see how the cartdrige distortion is affected, somehow.
The gain is somewhat self leveling by input Z of cartdrige, it seemed [lower output and lower Z cart... higher gain]. well, that was the idea. i dont remember much more about it anymore. I hope this info helps you in some way.
Enjoy !
Thx-RNMarsh
I didnt even know this subject existed until doing search on the Web about measuring Rbb and saw this .....
i owned a Denon DL103 so that is what i used. Medium output voltage.
I just had access
)) to those transistors I used - so i used them. No attempt to make the worlds lowest noise MC step-up but it was very quiet in the home music system. I had a Denon step-up transformer to compare against for noise and distortion and response. The important point I wanted to find out was -- if a lower than 100 Ohm cart load (the typical MC value used) made the sound have lower distortion as was reported via listening by Peter Moncrief, then could I make the input Z of a MC pre-preamp with a very low z and would it sound even better -- or at least better than what I had been using. What would it sound like to My ears. The passive Rload on the cart had been to make it sound 'cleaner' and lower distortion but, of course the cart output level was reduced as well. So it was just an interesting affect but not practicle.
It was developed to provide a very low input Z, direct coupled as much as possible and a sysmetrical design with the benefit of lower distortion coming off the cartrdige.
--- now that i dont even have a phono/LP any more --- to maximize its potential with better transistors etal. I gave it to Kavi, also. If I had a phono system I would love to hear what improvements you will make to it.
Eventually - try to see how the cartdrige distortion is affected, somehow.
The gain is somewhat self leveling by input Z of cartdrige, it seemed [lower output and lower Z cart... higher gain]. well, that was the idea. i dont remember much more about it anymore. I hope this info helps you in some way.
Enjoy !
Thx-RNMarsh
Last edited:
Has anyone put this little circuit to the SIM yet?
might be good - might be bad. Erno's method should be fine I would imagine. I had the JC MC (battery powerd) pre-preamp and a Denon transformer that is made for the Denon 103 cart and could compare but could not measure things like s/n and distortion at MC levels back then. At that time, I had a high efficiency custom made 3-way all JBL system and there was no noise heard (all horn... bass... mid..highs). Triamped with custom variable crossover and seperate amps. It was quiet and sounded the best of the three. I cant remeber any tech details this far beyond it.
Still looking for a follow-up on how much the distortion from the cartridge is reduced at the input to the amp via low Z loading.
THx-RNMarsh
might be good - might be bad. Erno's method should be fine I would imagine. I had the JC MC (battery powerd) pre-preamp and a Denon transformer that is made for the Denon 103 cart and could compare but could not measure things like s/n and distortion at MC levels back then. At that time, I had a high efficiency custom made 3-way all JBL system and there was no noise heard (all horn... bass... mid..highs). Triamped with custom variable crossover and seperate amps. It was quiet and sounded the best of the three. I cant remeber any tech details this far beyond it.
Still looking for a follow-up on how much the distortion from the cartridge is reduced at the input to the amp via low Z loading.
THx-RNMarsh
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.