How about sharing a schematic with us
I've been looking all over for a CCS conected 6C45 pi schematic, would you share this with us.
I've been looking all over for a CCS conected 6C45 pi schematic, would you share this with us.
Mike,
For a good CCS design, you should try check out Bas Horneman's website. You can use it on 6C45pi, and it works very good.
What application would you be using it for?
Regards,
Kevin
For a good CCS design, you should try check out Bas Horneman's website. You can use it on 6C45pi, and it works very good.
What application would you be using it for?
Regards,
Kevin
MIKET,
Nothing special to draw a schematic:
First of all, choose a CCS circuit to build. I have built Garry Pimm' s BBMCCCS. As Kevin suggests, there is also Bas Horneman' s CCS design, which is much simpler than the BBMCCCS, but you must use exclusively the depletion mode MOSFETs DN2540N5, which I had not.
After you have completed the CCS, you just connect it in place of the load resistor (or choke) of your circuit.
For the circuit I used it (6C45P in common cathode configuration), I set current at 18 mA and chose cathode resistor so that grid is biased at aprox. -1.2 V. With these values, plate voltage sets at around 120 V.
Regards,
Evangelos
Nothing special to draw a schematic:
First of all, choose a CCS circuit to build. I have built Garry Pimm' s BBMCCCS. As Kevin suggests, there is also Bas Horneman' s CCS design, which is much simpler than the BBMCCCS, but you must use exclusively the depletion mode MOSFETs DN2540N5, which I had not.
After you have completed the CCS, you just connect it in place of the load resistor (or choke) of your circuit.
For the circuit I used it (6C45P in common cathode configuration), I set current at 18 mA and chose cathode resistor so that grid is biased at aprox. -1.2 V. With these values, plate voltage sets at around 120 V.
Regards,
Evangelos
Hi,
That's were it should be...didn't you have trouble with another set of 6C45Pis?
Has anyone tried out the IXYS?
From what I've read it should be really easy to use.
Cheers.😉
That's were it should be...didn't you have trouble with another set of 6C45Pis?
Has anyone tried out the IXYS?
From what I've read it should be really easy to use.
Cheers.😉
kenev said:
Not quite. If you have a circuit, and you sub the CCS for a resistor, your dissipation in the CCS will likely be high because of excessive voltage drop across it. If you sub for a choke, then you will run out of voltage swing because a choke can output a signal above the B+ rail, but a CCS can't.
For a preamp, you'll need say 20V across the CCS (for most types), plus twice your expected max peak voltage swing, plus a bit more, say 20% "just in case". A CCS will hard clip unlike a resistive or choke loaded stage, so you need to make sure it never can. I also like sand to run cool, so be excessive with the heatsinking.
Heatsinking
Most CCSs end up with a power device screwed to a heatsink. Invariably, the output terminal of the CCS is connected to the conductive plate on the back of the device (either collector or drain). Mount this device on your earthed chassis with an insulating washer and you've just wrecked the HF performance of the CCS because of the shunt capacitance formed between the device and the chassis. This is one of the few instances where a little stand-alone heatsink is better than using the chassis!
Most CCSs end up with a power device screwed to a heatsink. Invariably, the output terminal of the CCS is connected to the conductive plate on the back of the device (either collector or drain). Mount this device on your earthed chassis with an insulating washer and you've just wrecked the HF performance of the CCS because of the shunt capacitance formed between the device and the chassis. This is one of the few instances where a little stand-alone heatsink is better than using the chassis!
Thanks EC for adding that, you are of course correct again.
I also wonder if that could be the reason why some people haven't liked them in some examples. I've only ever used small seperate heatsinks for each device in CCS applications.
I also wonder if that could be the reason why some people haven't liked them in some examples. I've only ever used small seperate heatsinks for each device in CCS applications.
Resistance is useless
It's just occurred to me that not only do you get the capacitance (about 10pF using an MJE350), but you also get the leakage resistance of the insulating washer. I couldn't measure the leakage resistance, but given that it's not difficult to make a CCS with an output resistance of 10M, almost anything could cause it to deteriorate, possibly in a non-linear way. Thoughts?
It's just occurred to me that not only do you get the capacitance (about 10pF using an MJE350), but you also get the leakage resistance of the insulating washer. I couldn't measure the leakage resistance, but given that it's not difficult to make a CCS with an output resistance of 10M, almost anything could cause it to deteriorate, possibly in a non-linear way. Thoughts?
Hi,
Frank,
Yes, indeed. This is the "correct" set. With the other one, plate voltage sets at around 150 V.
Which IXYS are you referring to? Is it a depletion mode MOSFET?
Brett,
Thanks for the correction - please delete the word "choke".
I leave enough voltage headroom for the CCS and also the MOSFET are adequately heatsinked (I can easily touch the heatsinks after 2 or 3 hours of use.
Regards,
Evangelos
Frank,
Yes, indeed. This is the "correct" set. With the other one, plate voltage sets at around 150 V.
Which IXYS are you referring to? Is it a depletion mode MOSFET?
Brett,
Thanks for the correction - please delete the word "choke".
I leave enough voltage headroom for the CCS and also the MOSFET are adequately heatsinked (I can easily touch the heatsinks after 2 or 3 hours of use.
Regards,
Evangelos
OK, just had another (probably) crazy idea.
I have 2 ECC99's and 2 CCS boards. Could I parallel the triodes in each ECC99 and set the CCS for 44mA? Would that work? Would it have any advantages (or disadvantages) over using a single ECC99 triode half? Would it be better with 4 CCS boards (set to 22mA each), one per triode? What about cathode bias resistors - shared or one per triode?
The obvious advantage and disadvantage of paralleling tubes that I'm aware of is decreased output impedance and increased Miller capacitance. Other than that, would I run into any 'sharing' issues between the two triodes, where they won't work well with each other? I've read of people using individual grid stoppers when paralleling tubes - would that help? What does that do anyway?
Saurav
I have 2 ECC99's and 2 CCS boards. Could I parallel the triodes in each ECC99 and set the CCS for 44mA? Would that work? Would it have any advantages (or disadvantages) over using a single ECC99 triode half? Would it be better with 4 CCS boards (set to 22mA each), one per triode? What about cathode bias resistors - shared or one per triode?
The obvious advantage and disadvantage of paralleling tubes that I'm aware of is decreased output impedance and increased Miller capacitance. Other than that, would I run into any 'sharing' issues between the two triodes, where they won't work well with each other? I've read of people using individual grid stoppers when paralleling tubes - would that help? What does that do anyway?
Saurav
Saurav,
The ECC99 has a ton of Miller C now, and paralleling would double it. Your Zout will be low enough as it is. Just use one side per channel. If using monoblocks, use triode A in one amp and triode B in the other, and swap the tubes over every so often. Ground the unused grid, and put a small current through the unused triode section.
The ECC99 has a ton of Miller C now, and paralleling would double it. Your Zout will be low enough as it is. Just use one side per channel. If using monoblocks, use triode A in one amp and triode B in the other, and swap the tubes over every so often. Ground the unused grid, and put a small current through the unused triode section.
Thanks.
> If using monoblocks, use triode A in one amp and triode B in the other, and swap the tubes over every so often.
It's a stereo amp, but I was planning to do that anyway.
> Ground the unused grid, and put a small current through the unused triode section.
That's new, never seen that advice before. Shouldn't be too hard to implement.
> If using monoblocks, use triode A in one amp and triode B in the other, and swap the tubes over every so often.
It's a stereo amp, but I was planning to do that anyway.
> Ground the unused grid, and put a small current through the unused triode section.
That's new, never seen that advice before. Shouldn't be too hard to implement.
Is -70dB crosstalk at 20kHz significant?
The ECC99 is very similar to the E182CC, for which Ca1-g2 = 0.1pF. If you had a source resistance of 25k, that 0.1pF would cause crosstalk at -70dB at 20kHz (less at lower frequencies).
I'd just use a single ECC99 - one half for one stereo channel, one half for the other.
The ECC99 is very similar to the E182CC, for which Ca1-g2 = 0.1pF. If you had a source resistance of 25k, that 0.1pF would cause crosstalk at -70dB at 20kHz (less at lower frequencies).
I'd just use a single ECC99 - one half for one stereo channel, one half for the other.
It took me a while to wrap my head around that one, I thought you were talking about Miller capacitance. Ca1-g2 is crosstalk between the triode halves, not Miller C.
This driver will be going into an amp which currently has SRPP 6SL7s. So I already have 2 tubes (and 2 holes in the chassis), which is why I bought 2 ECC99s and 2 "hole shrinkers" (things which allow 9-pin sockets to be put into octal holes), just to maintain the visual symmetry and not have an open hole in the front of my chassis.
This driver will be going into an amp which currently has SRPP 6SL7s. So I already have 2 tubes (and 2 holes in the chassis), which is why I bought 2 ECC99s and 2 "hole shrinkers" (things which allow 9-pin sockets to be put into octal holes), just to maintain the visual symmetry and not have an open hole in the front of my chassis.
Saurav,
Thanks for all these dumba$$ questions, this is a thread from which I will learn a LOT. I'm at page 1 now, and I'm going to grab my notebook, too much to remember right away. Thanks 🙂
hey-Hey!!!,
If you have Gary Pimm-style MOSFET cascode CCS you can significantly lower output Z by taking output from the lower FET's source instead of the other side of the current setting resistor...at least half an order of magnitude reduction v. single plate output Z at your operating point.
cheers,
Douglas
If you have Gary Pimm-style MOSFET cascode CCS you can significantly lower output Z by taking output from the lower FET's source instead of the other side of the current setting resistor...at least half an order of magnitude reduction v. single plate output Z at your operating point.
cheers,
Douglas
- Status
- Not open for further replies.