Single Bottle of 5670 Mic Preamp
- Thread starter Adky
- Start date
Help Support GroupDIY Audio Forum:
pstamler
Well-known member
I'm afraid there are going to be some loading and distortion issues with that design.
First, the loading. Most 150:15k transformers (1:10 winding ratio) want to see 150k at the secondary for optimum damping. By loading this one with 50k, there will be two effects. The transformer's frequency response may suffer by being overdamped, and you may get a rolloff at the top. Meanwhile, the input impedance will be 500 ohms (50k / 10^2). An SM57 would probably be pretty happy about that, but most condenser mics would not.
Second, take a look at the output. Assuming you load it wilh 600 ohms, and want to drive the output to +20dBu (that's 16dB over the standard +4dBu operating level, not a whole lot of headroom, but usually enough). +20dBu is 7.75Vrms, or about 10.96V peak. Driving that into 600 ohms requires about 18.3mA, so you'll have to operate the output tube at at least that current.
But...18.3mA across a 22k resistor drops a bit over 400V, and the plate supply is only 180V. So that's not going to fly; A resistor of 4.9k would drop about 90V at 18.3mA, leaving the cathode about halfway between ground and B+. It'll dissipate about 1.6W, so a 3W or bigger resistor would do the job.
But...at maximum output the tube needs to be swinging to about 36.6mA. How linear is the tube's behavior at current swings of zero to 36mA? My guess is, not very; the distortion will be high.
It's probably feasible to redesign the preamp for a 10k:600 ohm output transformer. That'd need only about 5mA of idle current (well, a bit more for linearitly's sake, but let's ignore that for the moment). But you'll need to crank up the B+, since +20dBu at the output means about +32dBu at the output transformer's primary. Run it at +300V, juggle resistor values, and you'll be in better shape.
Peace,
Paul
First, the loading. Most 150:15k transformers (1:10 winding ratio) want to see 150k at the secondary for optimum damping. By loading this one with 50k, there will be two effects. The transformer's frequency response may suffer by being overdamped, and you may get a rolloff at the top. Meanwhile, the input impedance will be 500 ohms (50k / 10^2). An SM57 would probably be pretty happy about that, but most condenser mics would not.
Second, take a look at the output. Assuming you load it wilh 600 ohms, and want to drive the output to +20dBu (that's 16dB over the standard +4dBu operating level, not a whole lot of headroom, but usually enough). +20dBu is 7.75Vrms, or about 10.96V peak. Driving that into 600 ohms requires about 18.3mA, so you'll have to operate the output tube at at least that current.
But...18.3mA across a 22k resistor drops a bit over 400V, and the plate supply is only 180V. So that's not going to fly; A resistor of 4.9k would drop about 90V at 18.3mA, leaving the cathode about halfway between ground and B+. It'll dissipate about 1.6W, so a 3W or bigger resistor would do the job.
But...at maximum output the tube needs to be swinging to about 36.6mA. How linear is the tube's behavior at current swings of zero to 36mA? My guess is, not very; the distortion will be high.
It's probably feasible to redesign the preamp for a 10k:600 ohm output transformer. That'd need only about 5mA of idle current (well, a bit more for linearitly's sake, but let's ignore that for the moment). But you'll need to crank up the B+, since +20dBu at the output means about +32dBu at the output transformer's primary. Run it at +300V, juggle resistor values, and you'll be in better shape.
Peace,
Paul
Hum...... 300Volt is @ the Max of a 5670 or 2C51 can handle, also could sub and move the 50K ohm pot to 100K ohm at the end of it to become the output control........ btw: after some listening test , I don't notice much distortion from it yet ...........
> 300Volt is @ the Max of a 5670
Yes, but the tube does not see the full supply voltage.
I estimate that at 180V supply the first tube plate is near 105V, the second tube has about 70V plate-cathode. So the supply voltage could be 500V(!) without exceeding the tubes' plate-cathode voltage ratings; however the heater-cathode and plate dissipation ratings will probably be exceeded. The 1.5W dissipation will be exceeded around 300V B+; the heater-cathode rating is already at the 100V limit with 180V B+, you should bias-up the heater to maybe +50V before you go to 250V-300V supply.
At 180V supply: The first tube idles at 4mA, the second near 5mA.
Gain of first tube is 21. Gm of second tube is under 2,000uMho, so cathode impedance is over 500 ohms. If loaded in 600:600 to 600r load (unusual today) then second stage gain is about 0.5. With modern 22K loading, gain will be over 0.8, limited by the fact that a "600" transformer is not real hi-Z.
Maximum output in hi-Z is almost 70V peak to peak or 25V RMS. With gain of 17 from first grid, input grid overload is 1.4Vrms; with 1:10 input transformer about 0.14Vrms at "150" input. With high impedance loading, it seems to be "OK".
Maximum output in 600r is like 5V p-p or 1.7V rms or +7dBm. Line-up level used to be +8dBm; this won't line-up. 600r systems today are usually +4dBm line-up with peaks 14db to 20dB higher; this plan will clip severely at normal old-fashioned line levels. And the THD will be high, like 5% at +6dBm and 1% at zero dBm.
With 10K load it can do 60Vp-p; a 10:600 transformation gives 15Vp-p or 5Vrms or +16dBm. However gain is now just 5 from first grid, 50 from "150" input. That may be ample for some work, insufficient for small choir in large church. (Classic tube console mike preamps often had gain as low as 35dB or 56, but they were followed by mix-network boosters and then line drivers, rarely used alone.)
REALLY good performance from a twin-triode needs a custom output transformer: one that can take DC and has a high primary impedance.
Oh: at 600r load and this tube, that 2.2uFd cap gives roughly 72Hz bass-cut. It may do better as the OT inductance resonates the cap; in fact it could have a large bass-bump around 50Hz. Which may be nice for some things, but isn't "accurate".
> I don't notice much distortion from it yet ...........
Then be happy.
Yes, but the tube does not see the full supply voltage.
I estimate that at 180V supply the first tube plate is near 105V, the second tube has about 70V plate-cathode. So the supply voltage could be 500V(!) without exceeding the tubes' plate-cathode voltage ratings; however the heater-cathode and plate dissipation ratings will probably be exceeded. The 1.5W dissipation will be exceeded around 300V B+; the heater-cathode rating is already at the 100V limit with 180V B+, you should bias-up the heater to maybe +50V before you go to 250V-300V supply.
At 180V supply: The first tube idles at 4mA, the second near 5mA.
Gain of first tube is 21. Gm of second tube is under 2,000uMho, so cathode impedance is over 500 ohms. If loaded in 600:600 to 600r load (unusual today) then second stage gain is about 0.5. With modern 22K loading, gain will be over 0.8, limited by the fact that a "600" transformer is not real hi-Z.
Maximum output in hi-Z is almost 70V peak to peak or 25V RMS. With gain of 17 from first grid, input grid overload is 1.4Vrms; with 1:10 input transformer about 0.14Vrms at "150" input. With high impedance loading, it seems to be "OK".
Maximum output in 600r is like 5V p-p or 1.7V rms or +7dBm. Line-up level used to be +8dBm; this won't line-up. 600r systems today are usually +4dBm line-up with peaks 14db to 20dB higher; this plan will clip severely at normal old-fashioned line levels. And the THD will be high, like 5% at +6dBm and 1% at zero dBm.
With 10K load it can do 60Vp-p; a 10:600 transformation gives 15Vp-p or 5Vrms or +16dBm. However gain is now just 5 from first grid, 50 from "150" input. That may be ample for some work, insufficient for small choir in large church. (Classic tube console mike preamps often had gain as low as 35dB or 56, but they were followed by mix-network boosters and then line drivers, rarely used alone.)
REALLY good performance from a twin-triode needs a custom output transformer: one that can take DC and has a high primary impedance.
Oh: at 600r load and this tube, that 2.2uFd cap gives roughly 72Hz bass-cut. It may do better as the OT inductance resonates the cap; in fact it could have a large bass-bump around 50Hz. Which may be nice for some things, but isn't "accurate".
> I don't notice much distortion from it yet ...........
Then be happy.
Actually, if this will only be loaded with 5K and up (typical modern inputs), I don't see a real problem with the output stage.
Changing to a 10K:10K OT with 2uFd cap and ~~500r tube is liable to give a very large bass-bump at 20Hz; the "mismatch" 600 ohm winding is likely to be fairly flat bass.
The input "mis-match" is notable but not severe: many mikes are fine with 500R-1K loading. And the low pot value means the electronic hiss hardly changes as gain is reduced: full-up you have 15K-25K source, half-up you have 16K source.
The 500r load on 200r mikes does sacrifice significant noise-floor: NF is maybe 2dB higher than theoretical, whereas an unloaded 150:15K transformer to a high current tube can do close to 1dB NF. In many studios, this is not important.
The input dynamic range, for hi-Z loading is 120dB (0.3uV noise to 300mV overload) which is very good. Sources from 300mV down to almost 10mV can be captured at 16 bit without corruption. 10mV with dynamic mike is 108dB SPL which is 93dB above typical noise floors in mike and good rooms.
Turn-down does directly raise the equivalent acoustic noise level. Instead of the pot, switch-out the cathode capacitor to get 6dB more input level with the same noise level. To go further, a simple 20dB pad ahead of the transformer is adequate.
Changing to a 10K:10K OT with 2uFd cap and ~~500r tube is liable to give a very large bass-bump at 20Hz; the "mismatch" 600 ohm winding is likely to be fairly flat bass.
The input "mis-match" is notable but not severe: many mikes are fine with 500R-1K loading. And the low pot value means the electronic hiss hardly changes as gain is reduced: full-up you have 15K-25K source, half-up you have 16K source.
The 500r load on 200r mikes does sacrifice significant noise-floor: NF is maybe 2dB higher than theoretical, whereas an unloaded 150:15K transformer to a high current tube can do close to 1dB NF. In many studios, this is not important.
The input dynamic range, for hi-Z loading is 120dB (0.3uV noise to 300mV overload) which is very good. Sources from 300mV down to almost 10mV can be captured at 16 bit without corruption. 10mV with dynamic mike is 108dB SPL which is 93dB above typical noise floors in mike and good rooms.
Turn-down does directly raise the equivalent acoustic noise level. Instead of the pot, switch-out the cathode capacitor to get 6dB more input level with the same noise level. To go further, a simple 20dB pad ahead of the transformer is adequate.
