I had some spare time on holidays so I build quick and dirty version of that improved AD829 clone that we discussed in post #3248. I know it's ain't pretty, but then again, I had worse looking protos, that made 433Mhz with 140dB noise floors. I didn't saved any thd measurements, as there was nothing to measure, so harmonics are masked with noise floor which is similar to starless. Also I thrown in some shunt supply's, they are a variation of salas V1, just optimized for higher olg, BF862 for Vref, also tight smd layout made possible usage of meaner mosfet's with higher gfs. I attached measured 50mA step response (shunt current set to 100ma, rise time limited to 200ns, upper trace output 2mv/div, lower trace load switching). That's with just 1uf poli smd for output capacitance, but I could throw anything I like at output (even low esr FM's) and it would stay stable. Noise was about 10uV, as I was hitting 5uV noise floor of my LNMP. I have to do more listening, but first impression is good. Sound is maybe a little bit too analytical for my taste, but details are well preserved. After all, it's opamp topology, just heavily biased for class A, so it has some sonic resemblance, but no opamp input stage that I heard would come close to this discrete thingy. I also liked balanced input, no signs of hum anywhere ! I don't think I take it any further than this, as I really need to make my self something usable to be able to enjoy my records again without soldering batteries for switching phono pre on 
So next week naked truth will have new aluminum box, new psu and ON switch.
So next week naked truth will have new aluminum box, new psu and ON switch.Attachments
That proto looks really cool !
Yes, with balanced the hum is gone. That is the major advantage.
When it sounds a bit analytic it can be that you have to readjust your cartridge.
I also found out that it is an advantage to have an adjustable subwoofer. A dB or two more can make a big difference. When i change the phono stage i usually have to readjust the woofer. I also use CD as reality check. For tonal balance comparison it can be very convenient. When you have the Naked Truth ready i would very much like to see the final circuit diagram.
Yes, with balanced the hum is gone. That is the major advantage.
When it sounds a bit analytic it can be that you have to readjust your cartridge.
I also found out that it is an advantage to have an adjustable subwoofer. A dB or two more can make a big difference. When i change the phono stage i usually have to readjust the woofer. I also use CD as reality check. For tonal balance comparison it can be very convenient. When you have the Naked Truth ready i would very much like to see the final circuit diagram.
Common guys, that proto is simply mad Joachim, of course I will post everything when done. Just couple more sanity checks with psu options, and maybe I should try cascoding j74/k170 with some bjt's (like in original schematics), but I'm not quite sure about possible benefits when running single lowish current pair. Well it's the way you did it, so there must be something behind this. I will try.
Joachim, of course I will post everything when done. Just couple more sanity checks with psu options, and maybe I should try cascoding j74/k170 with some bjt's (like in original schematics), but I'm not quite sure about possible benefits when running single lowish current pair. Well it's the way you did it, so there must be something behind this. I will try.
More pairs in parallel give you less noise into low impedances. You have to handle more current though. That makes cascoding with Fets difficult. Here the BJT cascode is better.
Also the input capacitance goes up so using MM may be a problem. You loose universality.
I do not care much about MM but some will.
Also the input capacitance goes up so using MM may be a problem. You loose universality.
I do not care much about MM but some will.
I have now measured 70 pairs of the BC327/337 that Frans has send me. Here the NPN go more astray in the higher regions. 50% though fall in the 410 to 450 range where they have matches in PNP. PNP culminates at 430 and NPN at 450-460.
6 trannis had really low Hfe so they fall on the way side. That is 4.29% of the population.
Conclusion is that it is no problem to find matches but the relatively high amount of under performers require selection. Can be fun though. It´s quite meditative.
6 trannis had really low Hfe so they fall on the way side. That is 4.29% of the population.
Conclusion is that it is no problem to find matches but the relatively high amount of under performers require selection. Can be fun though. It´s quite meditative.
Attachments
A very good year is what I can see for all of us
Hi Frans
I see a bzx84c6v2l zener from Vout to the base of the lower transistor in the constant voltage source that supplies the current source on the shunt.
What is it´s purpose ?
Hi Joachim.... I am reading and learning all the time nowadays
Hi Holger... long live and prosper
Hi Frans
I see a bzx84c6v2l zener from Vout to the base of the lower transistor in the constant voltage source that supplies the current source on the shunt.
What is it´s purpose ?
Hi Joachim.... I am reading and learning all the time nowadays
Hi Holger... long live and prosper
Last edited:
It is the voltage reference (offset from Vout) for the voltage source that feeds the CCS.