This page discusses the effects of tube aging on the RIAA filters and other amplifier stages ... there is a lot of discussion on tube aging today, both because the influence on tube aging on circuits like phono amps (or other stages with filters) is larger than normal, but also since all of us seem to e searching for those NOS - never been used before - tubes to use in our latest projects. And since everybody on eBay seems to have tube testers, and tube matching seems to be more important than making good designs in the first place, I decided to look into the effects of tube aging in my DIY projects.
I've doubted whether to put this page in the RIAA cookbook or in the tube background section. I've decided to put it here since it deals primarily with the physical effects of tube aging. When desiging your RIAA circuit it is good to keep this page in mind. Remember that it is nearly impossible to make a tube phono design that is always low noise and always correct in (passive) RIAA filter without spending massive amounts of money to cater for things you'll probably not hear. Sometimes your amp just needs some maintenance, and small tubes have a very long lifespan with good parameters.
Should you nevertheless be worried about anyway, than you can always build a solid state amp (which for phono is not a bad idea really), use lots of feedback or or or ...
And remember, I'm not the ultimate tube expert, I'm just getting the info I think is missing from lots of other places.
The circuit below is from Loekie, and there are several phono circuits alike in design. Some have ECC83/12AX7 as the first tube, others ECC88/6DJ8. But one thing in common: often the first stage of the amp is a CCS amplifier.

Especially for passive RIAA filters, the output impedance of the tube driving the filter network is part of the RIAA filter itself. Look for example to the circuit of Loekie above. As you can see, the output impedance Ru of a ECC83/12AX7 tube with 150k plate resistor and bypassed cathode resistor is 43k Ohms. Most components in the figure do have more or less constant values, except the values of Ru and Cmiller.
The last two are dependent on the tube and on the tube constans of these tubes. However tube constants are not really constant anyway. As tubes age, the cathode emission decreases over time.
Take the circuit above with the following values:
Rp = 150,000 Ohms
ra = 62,500 Ohms (plate resistance of preceding 12AX7 tube)
mu = 100 (12AX7 tube)
Rk = 2200 Ohms (but we'll bypass that for the moment with a large capacitor)
Using the output impedance calculation I calculated the following vaules in a Excel spreadsheet:
| Z_out new tube | Aging factor, increase in ra | Z_out as a result of aging | Change % in Z_out |
|
44,118 |
100% | 44,118 | 0% |
| 44,118 |
105% |
45,652 | 3.5% |
| 44,118 | 110% | 47,143 | 6.9% |
| 44,118 | 115% | 48,592 | 10.1% |
| 44,118 | 120% | 50,000 | 13.3% |
| 44,118 | 125% | 51,370 | 16.4% |
| 44,118 | 130% | 52,703 | 19.5% |
| 44,118 | 135% | 54,000 | 22.4% |
| 44,118 | 140% | 55,263 | 25.3% |
| 44,118 | 145% | 56,494 | 28.1% |
| 44,118 | 150% | 57,692 | 30.8% |
As you can see, these changes are significant. On the other hand, in our example the effective resistance for the RIAA filter changes from 44k1 + 115k = 159.1 k Ohms to 57k7 + 115k = 172k7 Ohms. The change for the RIAA filter from 159.1 to 172.5 k is 8.5 % for an increase in ra of the tube of 50%
< Calculate effect on RIAA curve >
Should we decide not bypass the cathode resistor, then the story is very much different. First of all, without bypassing there is less gain and there is a higher output impedance. But, this output impedance is less susceptible to tube aging. For example, the Z-out for new tubes is 98k2 Ohms without bypass, and for a ra increased with 50% its value for Z-out has increased to (only) 101k7 Ohms. An increase of only 3.5%.
In this configuration, the first resistor of the RIAA curve would have had the lower value of 159k1 - 98k2 = 60k9 This is an increase of 2.2%
The influence on the gain is much less and that's because its gain is dependent
on the relation between R_p and r_a and when r_a is small enough compared to
R_p, then the influence of tube aging on an CSS is not that large.
For other calculations, use the spreadsheet and see the effects of tube aging for your particular stage (Not all typical tube stages available yet in the spreadsheet !!)
There are several ways to help achieve correct RIAA filtering:
| Page 3: Tube Aging Examples >> |