I removed both side covers from one trafo. One side has nothing. The other side, hand written:
"DS050 600"
A long vertical line divides 0 and 6. I'll shoot an image for resale in case I replace these.
600 ohm drives my 9k ohm load reasonably well, yes?
Suppose I wanted a potential last line preamp and need a new trafo. How much better would be an entirely different circuit with a different tube or tubes?
Sum total required features: remote volume, 2 variable inputs + HT bypass fixed/unity gain. Suitable current for 9k ohm load.
"DS050 600"
A long vertical line divides 0 and 6. I'll shoot an image for resale in case I replace these.
600 ohm drives my 9k ohm load reasonably well, yes?
Suppose I wanted a potential last line preamp and need a new trafo. How much better would be an entirely different circuit with a different tube or tubes?
Sum total required features: remote volume, 2 variable inputs + HT bypass fixed/unity gain. Suitable current for 9k ohm load.
Last edited:
Yes, it can drive a 9k load.
If you turn the signal up all the way, the 5842 plate will be going to 2 x B+.
If the 5842 goes completely into cut-off, the plate voltage may go even higher.
Make sure that everything is hooked up and working at low volume (you can hear the music), or you may be tempted to turn the volume all the way up because you do not hear anything, and then it might suffer damage.
You might want to provide a load resistor on the output.
Perhaps something on the order of 1k to 2k Ohms (that should not significantly increase distortion, but might help prevent the above high voltage condition listed above).
What device are you driving that only has a 9k input impedance?
If you turn the signal up all the way, the 5842 plate will be going to 2 x B+.
If the 5842 goes completely into cut-off, the plate voltage may go even higher.
Make sure that everything is hooked up and working at low volume (you can hear the music), or you may be tempted to turn the volume all the way up because you do not hear anything, and then it might suffer damage.
You might want to provide a load resistor on the output.
Perhaps something on the order of 1k to 2k Ohms (that should not significantly increase distortion, but might help prevent the above high voltage condition listed above).
What device are you driving that only has a 9k input impedance?
Last edited:
Thanks very much. Sure, Re. output load R. What R and minimum W? Is OPT yellow > RCA pin un-inverted output? I thought all single tube gain preamps with coupling cap invert phase.
Main amp is ca. 2010 Electrocompaniet ECI 5 MkII integrated, 47k balanced (unbalanced unknown, presume 23.5k) + two Dayton SA1000 each 12k ohm (SA1000 in Dr. David Greisinger's phase quadrature tuning > 4 subs).
I use the ECI's HT input. To test the 5842 I shall start with its volume low-mid and same for the ECI's variable input. (ECI preamp is weak point.)
What is approximate 5842 gain?
A dual EL34 soon replaces the ECI.
Main amp is ca. 2010 Electrocompaniet ECI 5 MkII integrated, 47k balanced (unbalanced unknown, presume 23.5k) + two Dayton SA1000 each 12k ohm (SA1000 in Dr. David Greisinger's phase quadrature tuning > 4 subs).
I use the ECI's HT input. To test the 5842 I shall start with its volume low-mid and same for the ECI's variable input. (ECI preamp is weak point.)
What is approximate 5842 gain?
A dual EL34 soon replaces the ECI.
Last edited:
I am not familiar with any of your amplifier models.
But I see 23.5k and 12k Inputs.
How do you get 9k input?
The 5842 mu (u) is 43, and the plate impedance is 1.8k.
If you load the 5000 Ohm primary of the output transformer secondary with its rated tap impedance (600 Ohm load from 600 Ohm tap), then the gain is:
Tube gain = 5000/(5000 + 1800) x 43 = 31.6.
The transformer reduction in impedance is 5000/600 = 8.33
The reduction in voltage is Square Root of 8.33 = 2.87
31.6/2.87 = gain from the 5842 grid to the 600 Ohm output (loaded with 600 Ohm).
31.6/2.87 = 11.0
If you load the 600 Ohm tap with more than 600 Ohms, the mid frequency gain will be more than 11.
But, at extremely low frequencies, the transformer inductance comes into play, and the gain will be less than at midband.
And, at extremely high frequencies, where the distributed capacitance may be a factor, the gain will be less than at midband.
The frequency response at low frequency, and perhaps at high frequencies too, will be flatter if the output is loaded with 600 Ohms, versus a much higher load than 600 Ohms.
If the transformer has significant leakage reactance (not as likely with the 2.87 turns ratio), then the extreme high frequency can roll off if the load is 600 Ohms.
One thing that has not yet been talked about is the square wave response.
That is dependent on the transformer quality, and on the load on the secondary.
The plate impedance, rp, that drives the primary is pretty much fixed, for any reasonable current that the 5842 will be used at, so is not a variable.
But the transformer, and the load are the variables in this case.
A voltage gain of 11 is 20.8 dB.
But I see 23.5k and 12k Inputs.
How do you get 9k input?
The 5842 mu (u) is 43, and the plate impedance is 1.8k.
If you load the 5000 Ohm primary of the output transformer secondary with its rated tap impedance (600 Ohm load from 600 Ohm tap), then the gain is:
Tube gain = 5000/(5000 + 1800) x 43 = 31.6.
The transformer reduction in impedance is 5000/600 = 8.33
The reduction in voltage is Square Root of 8.33 = 2.87
31.6/2.87 = gain from the 5842 grid to the 600 Ohm output (loaded with 600 Ohm).
31.6/2.87 = 11.0
If you load the 600 Ohm tap with more than 600 Ohms, the mid frequency gain will be more than 11.
But, at extremely low frequencies, the transformer inductance comes into play, and the gain will be less than at midband.
And, at extremely high frequencies, where the distributed capacitance may be a factor, the gain will be less than at midband.
The frequency response at low frequency, and perhaps at high frequencies too, will be flatter if the output is loaded with 600 Ohms, versus a much higher load than 600 Ohms.
If the transformer has significant leakage reactance (not as likely with the 2.87 turns ratio), then the extreme high frequency can roll off if the load is 600 Ohms.
One thing that has not yet been talked about is the square wave response.
That is dependent on the transformer quality, and on the load on the secondary.
The plate impedance, rp, that drives the primary is pretty much fixed, for any reasonable current that the 5842 will be used at, so is not a variable.
But the transformer, and the load are the variables in this case.
A voltage gain of 11 is 20.8 dB.
Last edited:
Wow, thanks very much for your time and consideration.
I measured the ECI HT input unbalanced 49.6k ohm + 12k ohm sub amp = 9.66k ohm.
If I add a 640 ohm R calculator says net is 600.24 ohm. What is the load R minimum W?
21 dB is more than needed; I think my current preamp is 11.0 dB; on rarest occasion I could have used 1-2 dB more. Can I limit gain to 14 dB without degrading performance?
What transformer if any improves performance? Direct replacement?
What's your opinion of this tube circuit coupled with the right transformer?
I measured the ECI HT input unbalanced 49.6k ohm + 12k ohm sub amp = 9.66k ohm.
If I add a 640 ohm R calculator says net is 600.24 ohm. What is the load R minimum W?
21 dB is more than needed; I think my current preamp is 11.0 dB; on rarest occasion I could have used 1-2 dB more. Can I limit gain to 14 dB without degrading performance?
What transformer if any improves performance? Direct replacement?
What's your opinion of this tube circuit coupled with the right transformer?
Last edited:
Presuming this OPT has balanced outputs: is there any good reason for separate output to drive main and sub amps?
If yes, and if the two outputs are anti-phase, in a normal system I would swap sub output polarities. But Griesinger's phase quadrature runs L subs @ 45 degrees and R subs @ 135 degrees.
Unless I miss something inverting the bass 180 degrees makes no net difference in this sytem: it simply inverts sub phase L vs. R, but sub phase vs. mains is unchanged overall.
If yes, and if the two outputs are anti-phase, in a normal system I would swap sub output polarities. But Griesinger's phase quadrature runs L subs @ 45 degrees and R subs @ 135 degrees.
Unless I miss something inverting the bass 180 degrees makes no net difference in this sytem: it simply inverts sub phase L vs. R, but sub phase vs. mains is unchanged overall.
Your posted 'schematic', is technically only a wiring diagram.
I finally had to look at the pinout of the 5842, to be able to analyze the circuit.
The 5842 was originally designed to be used as a Grounded Grid RF amplifier.
That is why there are 4 grid pins, to reduce the lead inductance from the grid (and to connect all those pins to ground).
And for an RF grounded grid amp, that is also why the plate pin is opposite to the cathode pin.
I think you should only use 1 grid stopper resistor, and connect it to only 1 grid pin (just pick one . . . 4, 5, 7, or 8). Either use a 470 Ohm or a 1k Ohm, non-inductive resistor.
We are trying to stop RF oscillations, and stop RF amplification.
This is an Audio application.
You have many questions, and I do not seem to be able to keep up with all of them.
And they encompass things that are not about using/designing the 5842 preamp.
I hope somebody else chimes in for things related to subwoofer phase, etc.
As to a gain of 20.8 dB, that is across the 600 Ohm secondary.
If it is center tapped, and you ground the center tap, the gain from either leg to ground is 6 dB less. 20.8 - 6 = 14.8 dB, closer to what you seem to want.
And, the output impedance from one leg to ground is 1/4 of what it is across the complete secondary.
You can even put a 600 Ohm resistor from leg to leg, and still use the center tap for a ground. And yes, you get 2 outputs that way, in opposite phase.
In terms of using an output transformer on a preamp, I have never used one.
I have had lots of experience with power amplifier output transformers.
And, I have also used both single ended and push pull interstage transformers.
I no longer use interstage transformers.
I only use RC coupling.
I hope I have answered the amplifier questions you have had so far.
I do not understand your question: "If I add a 640 ohm R calculator says net is 600.24 ohm. What is the load R minimum W?"
Did you mean what wattage resistor to use to load the 600 Ohm secondary?
We simply need to calculate the maximum output power of your preamp.
And then I like for my resistors to be 3x that number.
For safety reasons, I like my power supply resistors to be 5x the power they dissipate.
I finally had to look at the pinout of the 5842, to be able to analyze the circuit.
The 5842 was originally designed to be used as a Grounded Grid RF amplifier.
That is why there are 4 grid pins, to reduce the lead inductance from the grid (and to connect all those pins to ground).
And for an RF grounded grid amp, that is also why the plate pin is opposite to the cathode pin.
I think you should only use 1 grid stopper resistor, and connect it to only 1 grid pin (just pick one . . . 4, 5, 7, or 8). Either use a 470 Ohm or a 1k Ohm, non-inductive resistor.
We are trying to stop RF oscillations, and stop RF amplification.
This is an Audio application.
You have many questions, and I do not seem to be able to keep up with all of them.
And they encompass things that are not about using/designing the 5842 preamp.
I hope somebody else chimes in for things related to subwoofer phase, etc.
As to a gain of 20.8 dB, that is across the 600 Ohm secondary.
If it is center tapped, and you ground the center tap, the gain from either leg to ground is 6 dB less. 20.8 - 6 = 14.8 dB, closer to what you seem to want.
And, the output impedance from one leg to ground is 1/4 of what it is across the complete secondary.
You can even put a 600 Ohm resistor from leg to leg, and still use the center tap for a ground. And yes, you get 2 outputs that way, in opposite phase.
In terms of using an output transformer on a preamp, I have never used one.
I have had lots of experience with power amplifier output transformers.
And, I have also used both single ended and push pull interstage transformers.
I no longer use interstage transformers.
I only use RC coupling.
I hope I have answered the amplifier questions you have had so far.
I do not understand your question: "If I add a 640 ohm R calculator says net is 600.24 ohm. What is the load R minimum W?"
Did you mean what wattage resistor to use to load the 600 Ohm secondary?
We simply need to calculate the maximum output power of your preamp.
And then I like for my resistors to be 3x that number.
For safety reasons, I like my power supply resistors to be 5x the power they dissipate.
Ah, cool, a custom set of transformers!
Gain will be somewhere around 16dB if you're using the entire secondary. Output impedance would be around 200 ohms.
Gain will be somewhere around 16dB if you're using the entire secondary. Output impedance would be around 200 ohms.
More on preamp output transformers
(very similar to interstage transformers).
Good ones are expen$ive
Single Ended ones are subject to hum pickup from power supply power transformers, power supply chokes, and if they are on a complete amplifier, are subject to signal pickup from the output transformers.
Single ended are most sensitive (preamp out, interstage, output transformers)
Push pull ones are less sensitive.
They are not only expen$ive, they need to be spaced from other transformers and chokes, and they need to be oriented 90 degrees different than the the other transformers and chokes.
They should not be mounted on magnetic steel chassis. That will conduct the magnetic fields no matter which orientation you use.
They all have a degree of limiting frequency response, phase, and also may appear to "ring" when they receive a square wave signal.
Generally, this is a not to worry, unless they are inside of a Global Negative Feedback Loop.
To repeat my earlier calculations, and add what effect loading has . . .
I think the gain is 20.8 dB if you load the secondary with 600 Ohms.
With a much higher load impedance, the gain is increased, but the frequency response suffers.
With a 200 Ohm load, the gain will be reduced, but the distortion will go up.
I hope we are moving along to a complete solution for you.
Design, build, listen, enjoy!
(very similar to interstage transformers).
Good ones are expen$ive
Single Ended ones are subject to hum pickup from power supply power transformers, power supply chokes, and if they are on a complete amplifier, are subject to signal pickup from the output transformers.
Single ended are most sensitive (preamp out, interstage, output transformers)
Push pull ones are less sensitive.
They are not only expen$ive, they need to be spaced from other transformers and chokes, and they need to be oriented 90 degrees different than the the other transformers and chokes.
They should not be mounted on magnetic steel chassis. That will conduct the magnetic fields no matter which orientation you use.
They all have a degree of limiting frequency response, phase, and also may appear to "ring" when they receive a square wave signal.
Generally, this is a not to worry, unless they are inside of a Global Negative Feedback Loop.
To repeat my earlier calculations, and add what effect loading has . . .
I think the gain is 20.8 dB if you load the secondary with 600 Ohms.
With a much higher load impedance, the gain is increased, but the frequency response suffers.
With a 200 Ohm load, the gain will be reduced, but the distortion will go up.
I hope we are moving along to a complete solution for you.
Design, build, listen, enjoy!
Last edited:
While the mu of a 5842 is 50, with a 5K load I would expect that to drop a fair amount, and it would depend a bit on the actual operating point as well. Loading the secondary down with a 1K resistor may not be such a bad idea.
With sensitive audio signals, keep them short as possible and screened if possible.
I would suggest screening is better than twisting.
I would suggest screening is better than twisting.
Yes. I was asking about the power rating for the 470k input shorting R and the 640 ohm output shorting R.
Earlier you rec. 600 ohm load for this OPT. The following 3 loads in parallel should be 600 ohm total, correct?
Main amp
Sub amp
Additional load R
The exact load impedance is not critical.
A 700 Ohm load, or a 500 Ohm load will not cause much difference in gain, distortion, frequency response, or other performance characteristics.
Power = (E rms(squared))/R
A 470k input resistor, with a 2.1Vrms CD player output (3.0V peak), will dissipate
9.4 microwatts (but only if the CD player is sending a steady full scale test tone).
I would probably use a 1/2 watt non inductive resistor there, not because of the power dissipated, but because my old hands and old eyes, do not like to use smaller parts.
Perhaps you meant the 470 Ohm grid stopper. I would use a 1/2 watt, again because smaller is difficult for me to handle and see.
if the gain of the preamp is 11, the 2.1Vrms from the CD player becomes 23.1Vrms.
Into a 600 Ohm load, that is 0.89 Watts, but only if the CD player puts out a steady full scale test tone.
A 2 or 3 watt non inductive resistor works there.
A 700 Ohm load, or a 500 Ohm load will not cause much difference in gain, distortion, frequency response, or other performance characteristics.
Power = (E rms(squared))/R
A 470k input resistor, with a 2.1Vrms CD player output (3.0V peak), will dissipate
9.4 microwatts (but only if the CD player is sending a steady full scale test tone).
I would probably use a 1/2 watt non inductive resistor there, not because of the power dissipated, but because my old hands and old eyes, do not like to use smaller parts.
Perhaps you meant the 470 Ohm grid stopper. I would use a 1/2 watt, again because smaller is difficult for me to handle and see.
if the gain of the preamp is 11, the 2.1Vrms from the CD player becomes 23.1Vrms.
Into a 600 Ohm load, that is 0.89 Watts, but only if the CD player puts out a steady full scale test tone.
A 2 or 3 watt non inductive resistor works there.
Last edited:
Sorry I omitted the tube diagram!
5842 has four grid pins 4, 5, 7, 8. My understanding then, is to:
Replace the two 470R with a wire short.
Replace the wire short between pin 4 and 5 with non-inductive 470R (or same between pins 7 and 8)
Do I have this right?
An earlier poster rec. volume pot pin 2 to ground via 470k ohm R.
Last edited:
Pick one, and only one, grid pin to connect one end of a single 470 Ohm grid stopper to.
The other end of the 470 Ohm grid stopper connects to the wiper of the Potentiometer volume control.
Do not wire the other 3 grid pins to anything, not to each other, not to a resistor . . . nothing connected to them.
The top end of the potentiometer connects to the inside connector of the RCA phono jack.
The bottom end of the potentiometer connects to ground.
The potentiometer wiper connects to the 'free end' of the 470 Ohm resistor (the grid end of the 470 Ohm is not the 'free end').
The outside connection of the RCA phono jack connects to ground (like at the bottom end of the potentiometer).
The other end of the 470 Ohm grid stopper connects to the wiper of the Potentiometer volume control.
Do not wire the other 3 grid pins to anything, not to each other, not to a resistor . . . nothing connected to them.
The top end of the potentiometer connects to the inside connector of the RCA phono jack.
The bottom end of the potentiometer connects to ground.
The potentiometer wiper connects to the 'free end' of the 470 Ohm resistor (the grid end of the 470 Ohm is not the 'free end').
The outside connection of the RCA phono jack connects to ground (like at the bottom end of the potentiometer).
At the GE tube manual's only operating conditions given for the 5842,
the plate resistance is 1800 Ohms, and the mu is 43.
5000/(5000 + 1800) x 43 = 31.6, the grid to plate gain.
The transformer 5k to 600 Ohm voltage reduction is Square root of (5000/600) = 2.89
31.6/2.89 = 10.9 (about 11).
That only applies with a 600 Ohm load on the secondary.
Gain of 11 is 20.8 dB, close enough.
Yes, at different operating voltages and currents, the tube mu and tube plate resistance will vary somewhat, but not very much for reasonable limits of voltage and current (where the tube is reasonably linear, such as not at or near cutoff).
the plate resistance is 1800 Ohms, and the mu is 43.
5000/(5000 + 1800) x 43 = 31.6, the grid to plate gain.
The transformer 5k to 600 Ohm voltage reduction is Square root of (5000/600) = 2.89
31.6/2.89 = 10.9 (about 11).
That only applies with a 600 Ohm load on the secondary.
Gain of 11 is 20.8 dB, close enough.
Yes, at different operating voltages and currents, the tube mu and tube plate resistance will vary somewhat, but not very much for reasonable limits of voltage and current (where the tube is reasonably linear, such as not at or near cutoff).
Last edited:
Thank you very much, esp. for being so clear on the wiper to tube connection. Volume wiper > 470 ohm .5W > any single grid pin 4, 5, 7, or 8 (remaining 3 grid pins open). Also volume wiper > 470k > ground (prevents potential tube damage @ maximum level IIRC).
I won't hear the preamp till Rs arrive.
If I understood your earlier post, all 3 secondary wires (black, orange, yellow) driving 2 RCAs (one non-inverted the other inverted) lowers the gain, which I prefer.
I'd be further indebted for your time if you explain the wiring to minimize gain, including # of shorting Rs (to minimize noise and distortion IIRC) and their location; and lastly which terminal(s) connect to chassis star ground.
(When received, the orange wire was open, yellow > RCA pin, black > RCA shield and chassis start ground.)
Main amp 48k load checked with ohm meter/power off. Sub amp maker specifies 12k load (with power off and ohm meter across RCA input, resistance started @ 12k and constantly increased over time).
I won't hear the preamp till Rs arrive.
If I understood your earlier post, all 3 secondary wires (black, orange, yellow) driving 2 RCAs (one non-inverted the other inverted) lowers the gain, which I prefer.
I'd be further indebted for your time if you explain the wiring to minimize gain, including # of shorting Rs (to minimize noise and distortion IIRC) and their location; and lastly which terminal(s) connect to chassis star ground.
(When received, the orange wire was open, yellow > RCA pin, black > RCA shield and chassis start ground.)
Main amp 48k load checked with ohm meter/power off. Sub amp maker specifies 12k load (with power off and ohm meter across RCA input, resistance started @ 12k and constantly increased over time).
Last edited:
OK, I reread post 27 and I think I got how to minimize gain 6 dB, ground the secondary center tap, and 600R across the two secondary legs.
I'll post a new diagram, and if any errors please let me know.
I'll post a new diagram, and if any errors please let me know.
With trafo primary shorted, using my not so great DMM, DCR of the three secondary conductors measure:
25.4R black to yellow
19.1R black to orange
6.3R orange to yellow
The above seems to indicate:
Sum total windings between black and yellow (when I received the preamp these 2 colors drove the output)
1/4 total windings between orange and yellow
3/4 total winding between black and orange
There appears to be no center tap. What am I missing?
What is gain if orange and yellow drive the output? Same Q for black and orange.
The preamp has two pair of RCA outputs. Again: I need to drive two sub amps each with 12k input, and main amp with 48k input.
If it helps performance, separate OPT taps can drive separate RCA outputs (for main and sub amps), and it's perfectly OK for the sub amp source to be inverted vs. the main amp source. If the preamp source has high frequency noise or FR aberration, I presume the LP SW filter eliminates it.
Thanks very much!
25.4R black to yellow
19.1R black to orange
6.3R orange to yellow
The above seems to indicate:
Sum total windings between black and yellow (when I received the preamp these 2 colors drove the output)
1/4 total windings between orange and yellow
3/4 total winding between black and orange
There appears to be no center tap. What am I missing?
What is gain if orange and yellow drive the output? Same Q for black and orange.
The preamp has two pair of RCA outputs. Again: I need to drive two sub amps each with 12k input, and main amp with 48k input.
If it helps performance, separate OPT taps can drive separate RCA outputs (for main and sub amps), and it's perfectly OK for the sub amp source to be inverted vs. the main amp source. If the preamp source has high frequency noise or FR aberration, I presume the LP SW filter eliminates it.
Thanks very much!
Last edited:
You can not be sure of the gain from the DCR numbers.
If the total secondary turns is a 600 Ohm output, then the gain of all the turns of the secondary winding is as we already calculated (11).
The windings stack up, one on top of the other, and that makes the wire longer, so more resistance per turn.
That makes the ratio of turns and ratio of DCR not exactly the same.
Estimated gain with a 600 Ohm load from black to yellow:
11 from black to yellow
8.3 from black to orange
2.7 from orange to yellow
You can do gains of 11 and 8.3 (ground black)
You can do gains of 8.3 and 2.7 (ground orange)
You can not do gains of 11 and 2.7 (can not ground 2 colors at once, it shorts the output windings).
If the total secondary turns is a 600 Ohm output, then the gain of all the turns of the secondary winding is as we already calculated (11).
The windings stack up, one on top of the other, and that makes the wire longer, so more resistance per turn.
That makes the ratio of turns and ratio of DCR not exactly the same.
Estimated gain with a 600 Ohm load from black to yellow:
11 from black to yellow
8.3 from black to orange
2.7 from orange to yellow
You can do gains of 11 and 8.3 (ground black)
You can do gains of 8.3 and 2.7 (ground orange)
You can not do gains of 11 and 2.7 (can not ground 2 colors at once, it shorts the output windings).
Last edited: