When I read this thread I wonder about one issue.
Hugh, and Rod are two persons here which are keen on using bootstrap for driving the output transistors.
I wonder what are the pros and cons?
Pros: More power with given design, more peak drive voltage, less parts, simplier power supply for the drive circuits.
Cons: Start-up thumps, poor bass performance(?), poor clipping characteristics(?)
Is the major reason to use bootstrap to save parts (=costs)?
Are there any sonic properties involved?
Hugh, and Rod are two persons here which are keen on using bootstrap for driving the output transistors.
I wonder what are the pros and cons?
Pros: More power with given design, more peak drive voltage, less parts, simplier power supply for the drive circuits.
Cons: Start-up thumps, poor bass performance(?), poor clipping characteristics(?)
Is the major reason to use bootstrap to save parts (=costs)?
Are there any sonic properties involved?
Hi P-A,
Nice post; useful, instructive.
More power, perhaps, but not strictly. While the VAS won't clip until several volts above rail, this is not true at the other rail, where the VAS emitter sits. Thus the 'more power' simply means clip is asymmetrical, which strictly means power at clip onset is no different to a VAS fed from a CCS.
Less parts, yes, certainly, but not many, and in any case a quality large bootstrap cap, being electrolytic and highly stressed, is quite expensive.
Simpler drive circuits: maybe; since the bootstrap is, by its nature, one sided; then clearly the VAS is single ended. This forces a single ended design, but locks out fully complementary designs. However, as the SET set will tell you, there's nothing wrong with single ended.......
Start-up thumps are infrasonic, that is, you don't hear them. Certainly they can move the cone of the driver slightly, but it's not audible and generally if you power the diff pair with a CCS even this thump is largely eliminated.
Poor bass performance is an interesting one. Because the low frequency performance of a bootstrap VAS generally drops off below 100Hz, the VAS starts to load up, varies its current more, and the OLG drops. This causes quite a bit of H2 artefact to appear in the VAS output, and together with the diminishing feedback factor this has two effects; the amp's output amplitude stays pretty much the same down to about 15Hz, but with increasing H2 content AND source impedance of the amp as a whole the bass wettens up. That is, it sounds richer, and very slightly looser, rather like a tube amp which suffers similar effects through the limited primary inductance of its output transformer.
When you do comparisons between a bootstrap amp and a CCS amp you notice these things clearly; I tried for years to get a CCS powered VAS to sound good, even creating a novel current source with fixed impedance to ameliorate drive variations at crossover in the output stage, but the bootstraps always sound better. This is particularly noticeable on Rock music, and interestingly chamber music, which seems to come alive with 'wetter' bass, in much the same way as this sort of music favours tubes.
Certainly not. The cost factors might introduce another $1 of parts; no more nor less. The reasons are subjective, and focus (as they should) on sound quality.
See above!
Cheers,
Hugh
Nice post; useful, instructive.
More power, perhaps, but not strictly. While the VAS won't clip until several volts above rail, this is not true at the other rail, where the VAS emitter sits. Thus the 'more power' simply means clip is asymmetrical, which strictly means power at clip onset is no different to a VAS fed from a CCS.
Less parts, yes, certainly, but not many, and in any case a quality large bootstrap cap, being electrolytic and highly stressed, is quite expensive.
Simpler drive circuits: maybe; since the bootstrap is, by its nature, one sided; then clearly the VAS is single ended. This forces a single ended design, but locks out fully complementary designs. However, as the SET set will tell you, there's nothing wrong with single ended.......
Start-up thumps are infrasonic, that is, you don't hear them. Certainly they can move the cone of the driver slightly, but it's not audible and generally if you power the diff pair with a CCS even this thump is largely eliminated.
Poor bass performance is an interesting one. Because the low frequency performance of a bootstrap VAS generally drops off below 100Hz, the VAS starts to load up, varies its current more, and the OLG drops. This causes quite a bit of H2 artefact to appear in the VAS output, and together with the diminishing feedback factor this has two effects; the amp's output amplitude stays pretty much the same down to about 15Hz, but with increasing H2 content AND source impedance of the amp as a whole the bass wettens up. That is, it sounds richer, and very slightly looser, rather like a tube amp which suffers similar effects through the limited primary inductance of its output transformer.
When you do comparisons between a bootstrap amp and a CCS amp you notice these things clearly; I tried for years to get a CCS powered VAS to sound good, even creating a novel current source with fixed impedance to ameliorate drive variations at crossover in the output stage, but the bootstraps always sound better. This is particularly noticeable on Rock music, and interestingly chamber music, which seems to come alive with 'wetter' bass, in much the same way as this sort of music favours tubes.
Certainly not. The cost factors might introduce another $1 of parts; no more nor less. The reasons are subjective, and focus (as they should) on sound quality.
See above!
Cheers,
Hugh
going between AKSA vs.
it mostly came from my experience.
Technically, I am not sophisitcated enough to answer that but I think most of the bootstrps use large electrolytic caps. they tend to become less cap-like at high frequencies. As such, the gain of the bootstrap goes down at high frequencies.
I am sure others can articulate a more precise answer but that's my understanding so far.
Technically, I am not sophisitcated enough to answer that but I think most of the bootstrps use large electrolytic caps. they tend to become less cap-like at high frequencies. As such, the gain of the bootstrap goes down at high frequencies.
I am sure others can articulate a more precise answer but that's my understanding so far.
Yes, a bootstrap is more stable at high frequencies.
The reason is one of the paradoxes of electronics, as positive feedback normally makes an amplifier less stable.......
With increasing frequency, and typically starting about 100KHz, the ESR of the bootstrap cap, being electrolytic, begins to rise. A large contribution to this is the self inductance of the foil of the cap; by Xl = 2pifL this steadily rises. As frequency rises further this higher ESR prevents ready charge flow to and from the cap, and by around 300KHz the capacitor has electrically 'shrunk' to a small portion of it's nominal capacitance.
This in turn begins to load up the collector of the VAS as the resistance to rail steadily diminishes. This pulls back gain.
In any voltage amplifier, we have poles, which define the point on the spectrum where negative feedback turns positive. In a global feedback loop we must bring OLG below unity by this pole frequency so that when feedback turns positive, gain is below one and it will NOT oscillate. This is normally implemented with a simple capacitor between collector and base of the VAS.
The bootstrap helps considerably in culling the gain to below unity by the pole frequency. This is a great strength because it means that there is no longer just the one 6dB/octave mechanism, lag compensation, for meeting the Bode-Nyquist criteria.
Cheers,
Hugh
The reason is one of the paradoxes of electronics, as positive feedback normally makes an amplifier less stable.......
With increasing frequency, and typically starting about 100KHz, the ESR of the bootstrap cap, being electrolytic, begins to rise. A large contribution to this is the self inductance of the foil of the cap; by Xl = 2pifL this steadily rises. As frequency rises further this higher ESR prevents ready charge flow to and from the cap, and by around 300KHz the capacitor has electrically 'shrunk' to a small portion of it's nominal capacitance.
This in turn begins to load up the collector of the VAS as the resistance to rail steadily diminishes. This pulls back gain.
In any voltage amplifier, we have poles, which define the point on the spectrum where negative feedback turns positive. In a global feedback loop we must bring OLG below unity by this pole frequency so that when feedback turns positive, gain is below one and it will NOT oscillate. This is normally implemented with a simple capacitor between collector and base of the VAS.
The bootstrap helps considerably in culling the gain to below unity by the pole frequency. This is a great strength because it means that there is no longer just the one 6dB/octave mechanism, lag compensation, for meeting the Bode-Nyquist criteria.
Cheers,
Hugh
peranders said:
I love this debate, because we have minds from all over the earth mulling over the problem. Look at the many flags of the various posters. This cross pollination of ideas is a very good thing.
Of course the answer to "why bootstrap?" is simply to increase the voltage swing of the VAS. We've already beat this to death here. I suspect those in the bootstrap camp will stay there, and likewise those in the current source camp will stay there too.
I've tried all of these approaches to increasing the drive swing and chosen the brute force method of using higher voltages to the front end. I generally use dual voltage doublers and shunt regulators for the front end supplies, which costs 6 caps, 4 rectifiers, 2 resistors and two zeners . Sounds like a lot, but actually fits in a small area and is not expensive compared to other alternatives, such as an outboard transformer. For the usual 10-15 ma front end current this is satisfactory with 1 mv ripple. This ends up working very well, since I get sufficient drive in both directions, symmetrical clipping behavior and so on. I get every bit of output power the output devices can safely and cleanly deliver.
With a bootstrap as commonly implemented, you have asymmetrical clipping, which seems like not a good thing, but actually since I don't like to listen to an amp clipping, it's not really an issue.
If you want to partially correct that asymmetry, you can put a bootstrap on the other side too, just an R/C from output to the front-end supply (Vas emitter). It will balance the overdrive somewhat. This was depicted in the 1970s RCA Transistor Manuals, so it's nothing new. For instance, there are about 5 examples in the 1975 book, typically using 50 mf/270 ohm. I have not tried this, just reporting that it was the vogue 30 years ago.
I would say that this is only partially true. Bootstrapping is enabling the VAS to deliver enough CURRENT (it is in fact something like cheating a CCS). Without bootstrapping the collector resistor of the VAS would have to be unreasonably small, leading to increased dissipation in the VAS transistor. This in turn would call for a VAS transistor with lower beta, fT and whatsoever.
Regards
Charles
Hello Hugh, where's my bootstrap
I am crazy to have my bootstrap to feel that outstanding sound some people told me.
Im my idea, can kick me with the boot, or with the strap, and i still crazy to put my hands over this wonder.
My friends is trying to corrupt me, they want to see schematics, they want to put on Ray X....and i said no!...Hugh do not want that!
- But Hugh will never know!
- But my conscience will know!
They are entirely crazy to our meeting to judgement.... we always make this as our fun.... we order to 30 years prison to some Harman Kardon.... we send to electrical chair some Pionner... we are really bad!...leach is already buried... today not more than sand!(the amplifier, long life to Mr. Leach!)
They saw your picture , and they say to me that you are in the middle of the ocean , pointing the skies, with .30 to destroy the airplane that is sending your AKSA 55 kits to damn Destroyer.
They said you are alike one old movie, one strabic (eyes convergence) German soldier trying to shot many airplanes he could see because the strabic eyes.... he said you do not put nothing in mail to me, that i loose money.. inside the pack is one stone!.... they are nervous with me , because i told them cannot see board circuit...only take a look with big distance and without turn it... and they told you made sandwiched boards with capacitors and resistor in the middle.
We are turning crazy (already i am) to see AKSA modules and to hear aksa sound.
Please, do not explode the airplane is bringing modules to me..I will not let them convince me.
No problems with customs, my wife is a graduate customs officer, and i have a lot of Federal Policeman Friends...no way to arrest my circuits, i think it is more easy that they bring here into my hands.
Have you heard about Lula?.... second degree cousin. hohohô
Carlos
I am crazy to have my bootstrap to feel that outstanding sound some people told me.
Im my idea, can kick me with the boot, or with the strap, and i still crazy to put my hands over this wonder.
My friends is trying to corrupt me, they want to see schematics, they want to put on Ray X....and i said no!...Hugh do not want that!
- But Hugh will never know!
- But my conscience will know!
They are entirely crazy to our meeting to judgement.... we always make this as our fun.... we order to 30 years prison to some Harman Kardon.... we send to electrical chair some Pionner... we are really bad!...leach is already buried... today not more than sand!(the amplifier, long life to Mr. Leach!)
They saw your picture , and they say to me that you are in the middle of the ocean , pointing the skies, with .30 to destroy the airplane that is sending your AKSA 55 kits to damn Destroyer.
They said you are alike one old movie, one strabic (eyes convergence) German soldier trying to shot many airplanes he could see because the strabic eyes.... he said you do not put nothing in mail to me, that i loose money.. inside the pack is one stone!.... they are nervous with me , because i told them cannot see board circuit...only take a look with big distance and without turn it... and they told you made sandwiched boards with capacitors and resistor in the middle.
We are turning crazy (already i am) to see AKSA modules and to hear aksa sound.
Please, do not explode the airplane is bringing modules to me..I will not let them convince me.
No problems with customs, my wife is a graduate customs officer, and i have a lot of Federal Policeman Friends...no way to arrest my circuits, i think it is more easy that they bring here into my hands.
Have you heard about Lula?.... second degree cousin. hohohô
Carlos