I have been so pleased with my Pass equip. that I am considering making this as a preamp for my Home theater- 6 channels. However, I would prefer not making six boards and grounding half of the circuit on each to make it SE.
Does anyone have any thoughts on the simplest solution? Just invert the signal switching hot and ground on the appropriate RCA connection and use both inputs for separate channels?
I was not sure if sharing a ccs mean that the circuits were coupled in a way that required differential operation only (or single ended with 1/2 grounded which is still differential) or if each half acted as independent inverting amps. I don't have much experience with this "X" stuff.
Any thoughts would be appreciated.
Does anyone have any thoughts on the simplest solution? Just invert the signal switching hot and ground on the appropriate RCA connection and use both inputs for separate channels?
I was not sure if sharing a ccs mean that the circuits were coupled in a way that required differential operation only (or single ended with 1/2 grounded which is still differential) or if each half acted as independent inverting amps. I don't have much experience with this "X" stuff.
Any thoughts would be appreciated.
Hi,
the outputs will allways have the same signal 180° out of phase so you can´t use one channel for two signals.
William
the outputs will allways have the same signal 180° out of phase so you can´t use one channel for two signals.
William
What am I missing? If the signal is always inverted, couldn't You just invert the signal coming in (or going out) so right + is inverted to right - by swapping the signal to ground and ground to signal in (or by something more horrible, like using an inverting opamp). So, we now have an inverted signal (180 degrees) that is fed into 1/2 a channel and exits, inverted again, in phase with original. In phase become out of phase, enters 1/2 BLS and then this out of phase becomes inverted by the original circuit, back to phase again. Just like if you used a inverted opamp and then had to incorporate a buffer to bring the signal back into phase- or you could just invert the outgoing cable instead of using the extra buffer.
My question really is, then, inverting or not, do both halves act as independent channels, inverting or not, like the Pass D1 output.
Or is there something about the circuit that would require that 1/2 the circiut is grounded- where it becomes basically an active ground? My method, on the other hand is a passive ground and probably inferior. I know that in the original Pass article for the BLS, Mr, Pass states that 1/2 of 1 channel can be grounded to make it SE. Why won't this other way work? Ignore the inverting part, assuming we can.
My question really is, then, inverting or not, do both halves act as independent channels, inverting or not, like the Pass D1 output.
Or is there something about the circuit that would require that 1/2 the circiut is grounded- where it becomes basically an active ground? My method, on the other hand is a passive ground and probably inferior. I know that in the original Pass article for the BLS, Mr, Pass states that 1/2 of 1 channel can be grounded to make it SE. Why won't this other way work? Ignore the inverting part, assuming we can.
Hi,
they are not independent. The current through the amp is always the same. If you raise the current in one half it is reduced by the same amount in the other half. So you will always have the same output signal (180° out of phase) from both halves.
William
they are not independent. The current through the amp is always the same. If you raise the current in one half it is reduced by the same amount in the other half. So you will always have the same output signal (180° out of phase) from both halves.
William
I m sorry to be ignorant about this. How can the current be reduced in one channel? I could understand how it would be inverted, but that problem can be solved by inverting the signal before entering the preamp. As they share a current sink and are biased to the same point, wouldn't the current be the same?
Hi,
it can be reduced in one half (Not channel!) by a negative input signal. The current in the other half will then be bigger by the same amount.
William
it can be reduced in one half (Not channel!) by a negative input signal. The current in the other half will then be bigger by the same amount.
William
At idle, yes the currents will be the same within the matching tolerance. However with a signal take a look at what happens:
Say a positive going signal is applied to the + input. This will increase the current in that branch of the differential. Since it is fed with a CCS that limits current to ~40 mA then the other branch must reduce its current by the same amount. Less current means less voltage across teh load resistors, therefore the voltage on the Drain gets higher.
It works the same way when the input signal goes negative. The signal reduces the current in the driven half, increases the current in the undriven side, the drop across the load resistor is higher and the Drain (output) voltage goes lower.
Another way of looking at it is the signal which appears at the Source of the driven side is applied to the Source of the undriven side.
Back to your original question, it appears that you are asking how to drive balanced input amps with single ended gear.
The absolute simplest way would be to ground the unused input of your amp and drive the amp with your unbalanced signal. You'd lose some gain that way.
The BOSOZ, in its various incarnations can be used as a single ended to balanced converter. When a CCS is used it does it quite nicely. Again you'd ground the inverting input.
Other solutions would be to use a DRV134 or the AD line driver (P/N escapes me) or one of the newer chips like Twisted Pear is working with. The Twisted boards (or any IC solution) will be a lot smaller than 6 BOSOZ boards and give off a lot less heat.
If you are just looking for a Pass unbalanced line driver, see the original BOZ.
Say a positive going signal is applied to the + input. This will increase the current in that branch of the differential. Since it is fed with a CCS that limits current to ~40 mA then the other branch must reduce its current by the same amount. Less current means less voltage across teh load resistors, therefore the voltage on the Drain gets higher.
It works the same way when the input signal goes negative. The signal reduces the current in the driven half, increases the current in the undriven side, the drop across the load resistor is higher and the Drain (output) voltage goes lower.
Another way of looking at it is the signal which appears at the Source of the driven side is applied to the Source of the undriven side.
Back to your original question, it appears that you are asking how to drive balanced input amps with single ended gear.
The absolute simplest way would be to ground the unused input of your amp and drive the amp with your unbalanced signal. You'd lose some gain that way.
The BOSOZ, in its various incarnations can be used as a single ended to balanced converter. When a CCS is used it does it quite nicely. Again you'd ground the inverting input.
Other solutions would be to use a DRV134 or the AD line driver (P/N escapes me) or one of the newer chips like Twisted Pear is working with. The Twisted boards (or any IC solution) will be a lot smaller than 6 BOSOZ boards and give off a lot less heat.
If you are just looking for a Pass unbalanced line driver, see the original BOZ.
Could Terry's design be modified with IRFP044 output transistors, coupled with a center-tapped choke, etc. etc., and used like the ZV7-T?
For kicks, I'm studying ways to play with Nelson's ZV7-T. This sure seems close to what I'm trying to accomplish.
John Inlow🙂
For kicks, I'm studying ways to play with Nelson's ZV7-T. This sure seems close to what I'm trying to accomplish.
John Inlow🙂
John,
At first glance, I'd say the idea is feasible. Let me think about it some, reread the ZV7 article a couple of times, and I may be able to help you fine out the details. I'm just finishing up a DAC project, so I need something new to sink my teeth into.
Cheers, Terry
At first glance, I'd say the idea is feasible. Let me think about it some, reread the ZV7 article a couple of times, and I may be able to help you fine out the details. I'm just finishing up a DAC project, so I need something new to sink my teeth into.
Cheers, Terry
Thanks Terry🙂
John
044's may not have enough voltage capacity, there's always 244's. Also, this could be another great way to utilize Russ White's pcbs.
John
044's may not have enough voltage capacity, there's always 244's. Also, this could be another great way to utilize Russ White's pcbs.
And for those of use who have 110db 1w1m horns, three watts output would be adequate. Since the output devices can handle 36 watts, why not substitute R3 and R4 with a center tapped choke load, which will increase efficiency, then up the bias and tap into the IRF610's low Ciss for lower distortion?
The 610 does have low transconductance... can you get gain from an FET with figures this low?
Fun stuff 🙂
John
Edit: R11 and R12 could be replaced with trim-pots to adjust offset. P1 will have to be beefy, or removed altogether.
The 610 does have low transconductance... can you get gain from an FET with figures this low?
Fun stuff 🙂
John
Edit: R11 and R12 could be replaced with trim-pots to adjust offset. P1 will have to be beefy, or removed altogether.
John,
Got a chance to give your idea a good look over, and it looks not only feasible, but pretty straightforward. As long as we do some decent matching of the FET’s, R11 and R12 can stay as fixed resistors without incurring any disadvantage with respect to output DC offset. As you said, replace R3 and R4 with a choke load, but from there I think we can have a bit more fun. Or in other words, I’ve a few additional ideas of how to make this into a really nice low power amplifier.
First, replace Q1 and Q2 with the Lovoltech power JFET’s, and Q3, Q4 and Q5 with something like IRF240’s. This gives some nice headroom for upping the bias.
Also, from hindsight, let’s simplify the cascode a bit. Combine R5 and R6 into one resistor, replace Z6 and Z7 with one resistor (or potentiometer) and use this as a voltage divider that acts as the voltage reference for both cascode FET’s.
Seeing as we are building this as a power amp, we can remove P1 entirely. Similarly, to achieve the bias levels that are needed, we will need to replace P3 with a fixed value power resistor.
As to how much gain we can produce, I know that for the IRF610 we can definitely achieve in the neighborhood of 24dB and still retain a reasonable degree of the supersymmetric feedback benefit, by increasing the values of R15 and R16 to 200kohm. If you want to go with the JFET’s I’ll have to take a closer look at the spec’s, but I believe we should still meet a decent gain spec.
This could get very interesting!
Terry
Got a chance to give your idea a good look over, and it looks not only feasible, but pretty straightforward. As long as we do some decent matching of the FET’s, R11 and R12 can stay as fixed resistors without incurring any disadvantage with respect to output DC offset. As you said, replace R3 and R4 with a choke load, but from there I think we can have a bit more fun. Or in other words, I’ve a few additional ideas of how to make this into a really nice low power amplifier.
First, replace Q1 and Q2 with the Lovoltech power JFET’s, and Q3, Q4 and Q5 with something like IRF240’s. This gives some nice headroom for upping the bias.
Also, from hindsight, let’s simplify the cascode a bit. Combine R5 and R6 into one resistor, replace Z6 and Z7 with one resistor (or potentiometer) and use this as a voltage divider that acts as the voltage reference for both cascode FET’s.
Seeing as we are building this as a power amp, we can remove P1 entirely. Similarly, to achieve the bias levels that are needed, we will need to replace P3 with a fixed value power resistor.
As to how much gain we can produce, I know that for the IRF610 we can definitely achieve in the neighborhood of 24dB and still retain a reasonable degree of the supersymmetric feedback benefit, by increasing the values of R15 and R16 to 200kohm. If you want to go with the JFET’s I’ll have to take a closer look at the spec’s, but I believe we should still meet a decent gain spec.
This could get very interesting!
Terry
All right!!!
I love your ideas, Terry. Indeed, this project can become amazing. If we can keep the DC offset down with fixed resistors, then perhaps we can utilize Russ White's pcbs.
I like the idea of the employing the lovoltech jfets. Is the footprint the same (I'm thinking of Russ' pcbs again) -- something to research.
Many thanks for your input,
John🙂
I just checked: the pinouts for the lovoltech are the same. cool...
Also, something worth mentioning: the lovoltech ZV9 doesn't speck out much better than the ZV7-T until the aleph current source is properly applied, then the ZV9 specs at three watts are just amazing. Perhaps the low Ciss of the 610 may better serve our needs?
I love your ideas, Terry. Indeed, this project can become amazing. If we can keep the DC offset down with fixed resistors, then perhaps we can utilize Russ White's pcbs.
I like the idea of the employing the lovoltech jfets. Is the footprint the same (I'm thinking of Russ' pcbs again) -- something to research.
Many thanks for your input,
John🙂
I just checked: the pinouts for the lovoltech are the same. cool...
Also, something worth mentioning: the lovoltech ZV9 doesn't speck out much better than the ZV7-T until the aleph current source is properly applied, then the ZV9 specs at three watts are just amazing. Perhaps the low Ciss of the 610 may better serve our needs?
Ok,
Something else just dawned on me: Nelson's ZV9's specs (without the aleph current source sensing resistors) had similar distortion specs as the ZV7-T.
But, and a rather large but at that, at three watts the ZV9 manages to match, or slightly exceed ZV7-T's specs while operating single ended. In a differential pair configuration, wouldn't one assume ZV9's distortion would be much lower?
I'm hoping so, since in effect this is what we're trying to do with the BOSOZ. This is very interesting!
John🙂
Something else just dawned on me: Nelson's ZV9's specs (without the aleph current source sensing resistors) had similar distortion specs as the ZV7-T.
But, and a rather large but at that, at three watts the ZV9 manages to match, or slightly exceed ZV7-T's specs while operating single ended. In a differential pair configuration, wouldn't one assume ZV9's distortion would be much lower?
I'm hoping so, since in effect this is what we're trying to do with the BOSOZ. This is very interesting!
John🙂
Gainclone
Hello,
Someone has tried the BOSOZ with Gainclone as power amplifier?
Impressions?
Thanks for your aid.
Alessandro
Hello,
Someone has tried the BOSOZ with Gainclone as power amplifier?
Impressions?
Thanks for your aid.
Alessandro
Yes, I have tried it with several different GC designs and it works very nicely with them, specially if they are a balanced design.
Cheers, Terry
Cheers, Terry
Hi Terry,
Looking at your schematics, I'm wondering if the ZVN3310 could be substituted for Q1 and Q2? I ask because the ZVP3310 (aforementioned counterpart) sounds so good in my ZV7-T.
John🙂
Looking at your schematics, I'm wondering if the ZVN3310 could be substituted for Q1 and Q2? I ask because the ZVP3310 (aforementioned counterpart) sounds so good in my ZV7-T.
John🙂
Q1 and Q2 each dissipate around 1/4W in my design, which is a little high for the ZVN3310's where longterm reliability is a concern (assuming I remember the power rating info correctly). However, there is nothing saying that the cascode couldn't be adjusted to lower the voltage across Q1,2 or for that matter reduce the bias current.
I agree that the ZVN and ZVP3310's are nice devices. I am currently working out the kinks out of a D1 style DAC I/V stage with a folded cascode using both ZVN's and ZVP's.
If you're serious about making a go with some changes to the circuit, let me know and I can suggest a few other alterations that I've considered making.
Cheers, Terry.
I agree that the ZVN and ZVP3310's are nice devices. I am currently working out the kinks out of a D1 style DAC I/V stage with a folded cascode using both ZVN's and ZVP's.
If you're serious about making a go with some changes to the circuit, let me know and I can suggest a few other alterations that I've considered making.
Cheers, Terry.
Oh, I'm very serious, Terry. I have one of Twisted Pair's stereo kits. It's been sitting around for the last six months waiting for something. I think that something is the ZVN3310.
This little fellow has such a sweet disposition! It's P counterpart made all the difference in my ZV7-T. Also, it has far more transconductance than the current fets being employed in the bzls, not to mention 1/3 the capacitance. What a wonderful gift from Nelson!
Any input would be greatly appreciated, cause I'm seriously knee deep in my ZV7-T and don't have the mental resources to adapt the bzls. Of course, if you're in the same boat with your projects, then the line stage can wait. It just seems like a great little project to chat about.
Best regards,
John🙂
This little fellow has such a sweet disposition! It's P counterpart made all the difference in my ZV7-T. Also, it has far more transconductance than the current fets being employed in the bzls, not to mention 1/3 the capacitance. What a wonderful gift from Nelson!
Any input would be greatly appreciated, cause I'm seriously knee deep in my ZV7-T and don't have the mental resources to adapt the bzls. Of course, if you're in the same boat with your projects, then the line stage can wait. It just seems like a great little project to chat about.
Best regards,
John🙂
- Status
- Not open for further replies.