Hi,
I have a EF86 preamp that I use with a ZenV4 and a GC. Since I moved to passive I/V conversion (TDA1543) I am having a very low sound, so I need more gain in the pre.
So I decided to add a E88CC in cathode follower mode after the EF86s. My problem is that simulation results show 150 V on the anode of the E88CC and I am in doubt if this is too much.
I tried the setup last night but for some unknown reason the heater of the E88CC refused to work. I am using a 12VDC supply and put a 12AX7 heater in series just to get the voltage drop. But the heater of the E88CC passed the cutrrent with only 1.5V across it, being around 10V across the heater of the 12AX7. Then I used a 6V psu and the E88CC worked fine
Miguel
I have a EF86 preamp that I use with a ZenV4 and a GC. Since I moved to passive I/V conversion (TDA1543) I am having a very low sound, so I need more gain in the pre.
So I decided to add a E88CC in cathode follower mode after the EF86s. My problem is that simulation results show 150 V on the anode of the E88CC and I am in doubt if this is too much.
I tried the setup last night but for some unknown reason the heater of the E88CC refused to work. I am using a 12VDC supply and put a 12AX7 heater in series just to get the voltage drop. But the heater of the E88CC passed the cutrrent with only 1.5V across it, being around 10V across the heater of the 12AX7. Then I used a 6V psu and the E88CC worked fine
Miguel
Attachments
The problem is:
The 12AX7 needs to have it's heaters in series, ie connect pins 4 and 5 for 300mA operation. It then expects to drop 12.6 volts.
The E88CC has only one heater option: 6.3 volts at 300mA, between pins 4and 5.
Clear as mud? Check the datasheets @
http://www.mif.pg.gda.pl/homepages/frank/vs.html
Cheers,
The 12AX7 needs to have it's heaters in series, ie connect pins 4 and 5 for 300mA operation. It then expects to drop 12.6 volts.
The E88CC has only one heater option: 6.3 volts at 300mA, between pins 4and 5.
Clear as mud? Check the datasheets @
http://www.mif.pg.gda.pl/homepages/frank/vs.html
Cheers,
Uno momento John...if the 12AX7 (ECC83) has it's heater connected in series then the current is 150ma at 12.6 volts, but connecting them in parallel will give you a consumption of 300ma at 6.3 volts.
Sooo...if Miguel connects the 12AX7 with pins 4 & 5 connected together as one terminal, and using pin 9 as the other, it can then be connected in series with the E88CC across the 12 volt DC supply.
Why the use of a 12AX7 just as a voltage dropper for a E88CC is another question...
Sooo...if Miguel connects the 12AX7 with pins 4 & 5 connected together as one terminal, and using pin 9 as the other, it can then be connected in series with the E88CC across the 12 volt DC supply.
Why the use of a 12AX7 just as a voltage dropper for a E88CC is another question...
> miguel2 has attached this image:
Aside from your heater-power troubles, the 330Ω cathode resistor is really too low for a line-level buffer.
A good value of load resistor for a cathode follower is much higher than the usual bias resistor.
There are two popular ways to do this:
The one on the left will self-bias, and can be adjusted without regard for the stage before it.
The one on the right saves several parts and a bass roll-off, but now you have to consider both stages together.
You see both ways in fancy designs.
(You can add a grid-stopper if you wish, though even for hot tubes like the 99 they are seen more often than they are needed.)
Any good high-output cathode follower tends to have large heater-cathode voltage. For line-level work you can usually hold this well below 100V, maybe around 50V, which is safe for most tubes. The left design can easily work that way, the right design requires tube X1 to sit with low plate voltage which may conflict with other goals.
Aside from your heater-power troubles, the 330Ω cathode resistor is really too low for a line-level buffer.
A good value of load resistor for a cathode follower is much higher than the usual bias resistor.
There are two popular ways to do this:
An externally hosted image should be here but it was not working when we last tested it.
The one on the left will self-bias, and can be adjusted without regard for the stage before it.
The one on the right saves several parts and a bass roll-off, but now you have to consider both stages together.
You see both ways in fancy designs.
(You can add a grid-stopper if you wish, though even for hot tubes like the 99 they are seen more often than they are needed.)
Any good high-output cathode follower tends to have large heater-cathode voltage. For line-level work you can usually hold this well below 100V, maybe around 50V, which is safe for most tubes. The left design can easily work that way, the right design requires tube X1 to sit with low plate voltage which may conflict with other goals.
miguel
u mention that you needed more gain from your pre-amp adding a cathode follower would not increase your gain. Cathode follower is like a current amp where it lower your output impedance by a factor of rp/mu+1 and your gain is
muRk/(rp+rk(mu+1)) where rp=tube plate resistance, rk= cathode resistor. Normally the gain is less than 1. How about running your first stage ef86 in pentode mode followed by a cathode follower. The layout you have drawn could be improve. In your cathode follower there is resistor at the anode. Normally it should not be used layout similar to PRR has drawn. When you use r2(1000K) in that configuration instead of PRR drawings the input impedance is normally of the resistor only but when connected in drawing one increases the input impedance by a factor of 7-8 times of r2. This helps in using a pentode first stage with high output impedance. Hope this helps.
u mention that you needed more gain from your pre-amp adding a cathode follower would not increase your gain. Cathode follower is like a current amp where it lower your output impedance by a factor of rp/mu+1 and your gain is
muRk/(rp+rk(mu+1)) where rp=tube plate resistance, rk= cathode resistor. Normally the gain is less than 1. How about running your first stage ef86 in pentode mode followed by a cathode follower. The layout you have drawn could be improve. In your cathode follower there is resistor at the anode. Normally it should not be used layout similar to PRR has drawn. When you use r2(1000K) in that configuration instead of PRR drawings the input impedance is normally of the resistor only but when connected in drawing one increases the input impedance by a factor of 7-8 times of r2. This helps in using a pentode first stage with high output impedance. Hope this helps.
Hi,
Let's not forget a few things,Nick:
Using a penthode as a voltage amplifier will have a noise penalty, not to mention the fact that's far less linear than a triode and the gain would be sky-high, quite likely overloading the ECC88 following it.
In triode mode the EF86 has roughly the same gain as the ECC88, which makes me wonder why he'd use it in the first place since a single ECC88 could easily take care of things.
Good EF86s ain't exactly cheap either...
Cheers,
Let's not forget a few things,Nick:
Using a penthode as a voltage amplifier will have a noise penalty, not to mention the fact that's far less linear than a triode and the gain would be sky-high, quite likely overloading the ECC88 following it.
In triode mode the EF86 has roughly the same gain as the ECC88, which makes me wonder why he'd use it in the first place since a single ECC88 could easily take care of things.
Good EF86s ain't exactly cheap either...
Cheers,
For a preamp it depends on the gain needed and whatever attenuation, tone or volume control circuits for instance, have to be pushed afterward; but anyway, Frankenhouse suffers no ill effects from using a pentode, and incidentially, a seemingly dirty type which used to be used in TVs (a 6U8 or 6GH8).
Tim "Ending anti-pentode and anti-12AU7 criticicism since 1898!" (to tie in with my sig)
Tim "Ending anti-pentode and anti-12AU7 criticicism since 1898!" (to tie in with my sig)
fdegrove said:
Um, nope, depends how you use them. The last version of my GM70 amp was running pentodes for drivers, as was the version before last of the KT88/6550 amp. The 6550 version was changed because it had seperate driver and output stage supplies and I needed some of the parts elsewhere. But it sounded awesome, and when I'm done with the other PSU, I'll put them back.
http://www.pmillett.addr.com/pentodes.htm
Basically using something very similar to the 12HG7 setup.
Oh, and I won't mention that my 'main' power amp is all pentode, operated as pentodes and it's by far the best amp I've ever heard.
Egads, I've agreed with Tim.......but 12AU7's still suck.
true
about pentodes being noisy i have never used them before in input stages. Every textbook says so about pentodes being noisy. BUt his design uses a pentode as the front stag,e only way to get higher gain while using the same setup up right? cost is cruical for me BUt when high gain is needed pentode can swing high amounts with some current also depending on the tube used. Thorsten Loerch uses a el84 to drive a 300b. He states and graphs the thd of the whole amp with various layouts srpp,2 stage, cascode and pentode. Pentode being the lowest thd. About the gain can't we used a voltage divider after the pentode of before the pentode to control the voltage out from the pentode preferably high resistance voltage divider resistor so not to load the pentode.
about pentodes being noisy i have never used them before in input stages. Every textbook says so about pentodes being noisy. BUt his design uses a pentode as the front stag,e only way to get higher gain while using the same setup up right? cost is cruical for me
BUt when high gain is needed pentode can swing high amounts with some current also depending on the tube used. Thorsten Loerch uses a el84 to drive a 300b. He states and graphs the thd of the whole amp with various layouts srpp,2 stage, cascode and pentode. Pentode being the lowest thd. About the gain can't we used a voltage divider after the pentode of before the pentode to control the voltage out from the pentode preferably high resistance voltage divider resistor so not to load the pentode.Hi,
Heater:
I only used one section of the 12AX7 heater, so the voltage drop should be 6.3 V. The problem was that it had a voltage drop of 10V, being the E88CC with the remaining 1-2V. When powered by a 6VDC psu the E88CC works fine. I still dont get it... The 12AX7 is only there while I am testing the thing, as for now I only have a 12VDC psu on this preamp.
Pre section:
I already had the EF86 section, which is based on a design published on Svetlana's site. The original design is on the picture. They advise an input impedance of over 200k on the power amp. So I guess that the EF86s are working quite hard to drive the ZenV4, and I guess that voltage swing is considerably lower with this low input impedance power amp. That is why I want to add a cathode follower stage. And I would like to try it first with a cathode follower, and then if needed I convert it to penthode operation.
I used a resistor in the anode of the E88CC - this should not be used? The datasheets of the E88CC mention an operation at 100V, so isnt 250V too much? And the cathode resistors of 330/10k work well?
Do I have to raise the heater voltage? How is this done?
BTW I am using old phillips EF86s and E88CC.
Miguel
Heater:
I only used one section of the 12AX7 heater, so the voltage drop should be 6.3 V. The problem was that it had a voltage drop of 10V, being the E88CC with the remaining 1-2V. When powered by a 6VDC psu the E88CC works fine. I still dont get it... The 12AX7 is only there while I am testing the thing, as for now I only have a 12VDC psu on this preamp.
Pre section:
I already had the EF86 section, which is based on a design published on Svetlana's site. The original design is on the picture. They advise an input impedance of over 200k on the power amp. So I guess that the EF86s are working quite hard to drive the ZenV4, and I guess that voltage swing is considerably lower with this low input impedance power amp. That is why I want to add a cathode follower stage. And I would like to try it first with a cathode follower, and then if needed I convert it to penthode operation.
I used a resistor in the anode of the E88CC - this should not be used? The datasheets of the E88CC mention an operation at 100V, so isnt 250V too much? And the cathode resistors of 330/10k work well?
Do I have to raise the heater voltage? How is this done?
BTW I am using old phillips EF86s and E88CC.
Miguel
Attachments
Because the screen is connected to the plate, we call that a triode-strapped pentode, or in a jif, just triode (as "EF86 in triode"). EF86 I think is usually considered pretty blameless, as pentodes go.
The reason you get 10V across your 12AX7 heater is because it has more resistance than the ECC88's heater. Much more. So it forms a voltage divider with more voltage across the 12AX7.
To solve series heaters, you have to go with the highest current tube, in this case the ECC88. It does 300mA (note: I usually see 365mA?) at 6.3V, and since Iin = Iout, all other heaters in the chain have to have the same current capacity. Simple enough with a 12AX7 since paralleling the heaters (as mentioned in other posts) produces a 6.3V 300mA heater. Add the voltage, 6.3 + 6.3 = 12.6V is how much you have to supply. Iin = Iout so the current draw is 300mA.
<Tutorial mode: on>
OTOH, what if you have tubes that don't have the same heater current? In this case, if you only had an 18V transformer instead of the 12.6 you have, you could connect the 12AX7 in series with the ECC88. But wait, the currents aren't the same! So you connect a resistor in parallel with the 12AX7 to shunt the remaining current past it! Wasteful but it works. In this case, the first section is 6.3V at 300mA -- leaving 18 - 6 = 12V at 300mA for the 12AX7. That resistor I mentioned, it has to carry 300 - 150 = 150mA. Well Ohm's law states R = V/I so it must be 12 / .15 = 80 ohms.
<Tutorial mode: off>
Tim
EF86 I think is usually considered pretty blameless, as pentodes go.The reason you get 10V
across your 12AX7 heater is because it has more resistance than the ECC88's heater. Much more. So it forms a voltage divider with more voltage across the 12AX7.To solve series heaters, you have to go with the highest current tube, in this case the ECC88. It does 300mA (note: I usually see 365mA?) at 6.3V, and since Iin = Iout, all other heaters in the chain have to have the same current capacity. Simple enough with a 12AX7 since paralleling the heaters (as mentioned in other posts) produces a 6.3V 300mA heater. Add the voltage, 6.3 + 6.3 = 12.6V is how much you have to supply. Iin = Iout so the current draw is 300mA.
<Tutorial mode: on>
OTOH, what if you have tubes that don't have the same heater current? In this case, if you only had an 18V transformer instead of the 12.6 you have, you could connect the 12AX7 in series with the ECC88. But wait, the currents aren't the same! So you connect a resistor in parallel with the 12AX7 to shunt the remaining current past it! Wasteful but it works. In this case, the first section is 6.3V at 300mA -- leaving 18 - 6 = 12V at 300mA for the 12AX7. That resistor I mentioned, it has to carry 300 - 150 = 150mA. Well Ohm's law states R = V/I so it must be 12 / .15 = 80 ohms.
<Tutorial mode: off>
Tim
Hi,
Fine, only the situation is a little different here.
The circuit is fed from the output of a DAC that has it's voltage amplification stage(s) removed.
That loss now needs to be jacked up by the preamp using the circuit as proposed by Miguel.
From that context I'd stick with a trioded EF86 or I'd revise the whole thing and go for a different approach alltogether.
Cheers,
Brett said:
Fine, only the situation is a little different here.
The circuit is fed from the output of a DAC that has it's voltage amplification stage(s) removed.
That loss now needs to be jacked up by the preamp using the circuit as proposed by Miguel.
From that context I'd stick with a trioded EF86 or I'd revise the whole thing and go for a different approach alltogether.
Cheers,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.