Posted by
Walt N. |
March 11, 2022 - 10:16 am | |
Little more info from testing. I have an external phono pre-amp that I used with line-in on my computer. It worked there. According to the Cortina schematic, it will just bypass the phono pre-amp for Tuner or Aux settings. So, I hooked up the external pre-amp to the tuner inputs and the turntable to the pre-amp. Results were about the same. I'm now starting to believe that the Shure V-15 cartridge in the turntable is just not making enough signal? In any case, I now have two phone pre-amps that are not giving me enough signal. Frustrating. |
Posted by
Walt N. |
March 10, 2022 - 07:54 pm | |
I have rearranged my stuff for various reasons. Now, the Cortina is connected to my EICO RP-100 tape deck that I restored. The RP was working just fine getting its signal from my BD player. However, I find now that I barely have enough signal coming from the Cortina to move the meters unless I jack them way up. The record level pots in the RP-100 are not perfect, but they were working fine before, so I suspect that for some reason I'm not getting enough signal from the RP-100. Signal is clean. This is from an LP so is coming from the Phono pre-amp. The amp seems to work just fine playing LPs, though. Any thoughts? Maybe I need to work on the pre-amp after all? |
Posted by
Steve L. |
December 20, 2021 - 07:57 pm | |
Hi Rob, I truly appreciate your thoughtful comments. Roughly, the reason that Q302 isn't in danger of thermal runaway is that Q4 and R14 act to stabilize thermal changes in Vbe of Q302. More generally, the sum of the Vbe's of Q303, Q2, Q302 plus the drops across R12 and R14 must match the sum of the drops across R30, CR301 and CR302. Let's call those last three items, the Bias Source and call the others, the load. If Vsrc is effectively increased by a reduction in Vbe of Q302, I would expect most of the increase would appear in the drops across R12 and R14. For example, for temperature-dependent voltage effects, they handle the fact that there is one less diode junction in Vsrc than in the load. The voltage drop across R311 is on the order of only 20mV and doesn't have much effect.
If you need a more detailed explanation, I would suggest setting up a Spice simulation of the circuit and introducing a perturbing voltage source in the source or load. You should see that most of the perturbation does indeed appear across R12 and R14. Since this is getting rather arcane for most readers, please contact me directly if you would like to discuss it further. My email address is on the About page which is shown when you click the "About..." button near the top of the home page. Thank you for the stimulating post! |
Posted by
RobS |
December 17, 2021 - 01:02 am | |
I forgot to mention that resistor can also helps with the symmetry of the output stage. There's some articles in 1969 Wireless World by Shaw and Baxandall which do this in a slightly better way. |
Posted by
RobS |
December 17, 2021 - 12:50 am | |
Nice write up.
Regarding the mystery 4.7 ohm resistor, http://tronola.com/html/cortina_3070_page-3.html
The reason it is there is to ensure the drivers don't get into a thermal runaway condition. Not unlike emitter resistors on the output stage.
Assume the bias diodes present a fixed bias voltage to the driver. The diodes don't sense the driver temperature. Now assume the drivers increase in temperature by dT. That causes the Vbe of the drivers to decrease. Since the "base to base" bias is fixed the Vbe decrease causes an increase in voltage across the driver emitter resistor. (In this case, there's two driver Vbe. ) That increases the driver current. The increase in driver current causes an increase in power dissipation (approx VCC * change in current). The increase in power dissipation causes the device to heat up according to dT' = Rth * increase in Pdis where Rth is the thermal resistance of the driver. If dT' ends up being greater than the initial dT get a thermal runaway situation. If you go through the math for VCC = 40V, Rth = 150 to 200C/W and a total 2*Vbe temperature coefficient of 4mV/degC we end-up with that the emitter resistor needs to be more than 24 to 32 ohm. If we allow for the change in VBE with current re = VT / IC we could shave off 15 ohms or so; but at some risk. The point is you end-up with quite a small value. Many quasi-complementary amps have these small resistor on the drivers. |
Posted by
Steve L. |
November 25, 2021 - 03:02 pm | |
My pleasure and your kind words are most appreciated. I agree about the power transistors. Those and the driver substitutes weren't intended to be upgrades (though they are to a certain extent). Bravo for regreasing them; that made a big difference in transistor temperature in my unit. |
|