What is this topology called?

[teemuk]

Attached is a link to the schematic of a Fender Steel King amplifier.

http://www.fender.com/support/amp_sc...Steel_King.pdf

The power amplifier uses some sort of "semi symmetric" topology that creates a suitable drive signal for the other half of the symmetric VAS arrangement by inverting the output signal of the differential. I'd like to know more about this circuit but searching for information is always a bit vain if you can't come up with a proper search term. Is there a name for this circuit? Is it patented? No help of Slone's and Self's books either: Although the first one does cover symmetric, "mirror image" circuits it only focuses in completely symmetric circuits.

The Fender amplifiers are the only ones I've seen using this circuit.

Anyone?

[AKSA]

Hi Teemuk,

This is also used on the BXR300, a very reliable 300W Fender amp.

It's a clever system of shifting level to the other rail to allow a single diff iput stage to drive a VAS from both rails. If I had to give it a name, I'd call it a 'Level Shift double VAS' configuration.

Q6 is driven from the identical operating point as Q10, the upper VAS. Q6's collector is loaded identically to Q4, the obverse LTP device. Thus drive is identical to Q10 at the base of Q11, although you will notice C25, added presumably to maintain stability. Q7 and Q8 are overload clippers.

A lot of lag compensation is used on this design, but it sounds pretty good. The idea is to throw the amp into manacles to keep stability in any load, as on the road it will have long cables and nasty parasitics spread around.

I have tested the level shift function in isolation on the bench and it is very fast; 6dB point around 1MHz. However, with all the phase inversions this is not a particularly straightforward design and after a lot of agonising I eventually dropped it to pursue a single ended VAS which is simpler, offers better speed, can be stabilised more easily, and gives superior offset control, even with resistive loading on the LTP.

But it sure is a very appealing circuit......

Cheers,

Hugh

[Ipanema]

Hi Hugh,

Thanks for your insightful explanation. :-)

I observe that voltage at TP9 and TP10 are difference by 2volts, which might cause unbalance load to the LTP output. Will adding three more series diodes at the collector of Q5 helping it to balance out the LTP load?

Thanks.

[unclejed613]

ibteresting..... Q6 is a current mirror. Q7,8 are an interesting implementation (along with D19,20, 21, 23,24,25,27, 28) of a "baker clamp", which keeps the drivers and output devices out of saturation at clipping. yes, i agree, 220pf is a LOT of compensation, but Q10,11 have an ft of 50Mhz, and it takes a lot to tame them.

[AKSA]

Hi Ipanema (where is that girl??),

Imbalance in a LTP is principally current imbalance, and it's determined by:

1. Hfe and Vbe match of diff pair devices
2. Bias impedance at both device bases.

If the devices are well matched, and the bias impedance at the base is the same, here 27K, then the VASes top and bottom will always be driven to ensure zero output offset wrt ground AT IDLE. The necessary situation to achieve this will be equal current through each arm of the LTP.

The input stage runs at 2mA (9.4V/4K7), so 1mA in each arm represents balance. If you do the sums on this circuit you do find it's in balance regardless, because the bias requirement of Q7, the Baker clamp which pulls the VASes out of clip (protecting expensive speakers and also the output stage), will only exist when the circuit is near clipping. That is, Q7 won't be on until the collector of the upper VAS, Q10, is two junctions higher in potential than TP9. Since TP9 is close to the rail, only around 3.25V off the rail, Q7 will remain unbiased until the amp approaches clip. This, in my view, is the cleverest part of the circuit, and absolutely mandatory protection on a pro amp.

Jed,

220pF is huge. I can very nicely compensate a single ended VAS with single LTP input using just 47pF, and that with a 100MHz VAS, so this very high compensation is a result of the level shift, antiphase drive and the onerous operating environment of a pro amp. The Steel King and BXRs almost never came in repair when I was working briefly in the pro industry; they are robust as hell. Of course, there is a price for this rock solid stability, and that is generally sound quality......

Cheers,

Hugh

[destroyer X]

Her daugther and her grandaugther ... they have not her beauty.

Well...she is an old lady.... but sadly not so beautifull as during the sixties and seventies.

Here she is...image from 2001

Sorry by off topic...bye!

regards,

Carlos

[unclejed613]

there are some well designed pro amps that sound good as well as being built like a tank. but then again, you get what you pay for. the higher end amps use protection relays, while the cheaper ones use the base drive clamps such as the ones in the fender amp above. unfortunately, base drive clamps sound terrible when they activate, and they don't always activate only when there's a short either. they tend to activate when the amp is seriously overdriven, and make the distortion from clipping worse. the idea of using a baker clamp is to prevent overhang from saturation, which also makes it difficult to hear clipping until you get above 2 to 5% THD, but if the protection clamps activate also, the THD will instantly go from 5% to 15 or 20%, making the clipping very obvious.

[ilimzn]

Actually, excluding the clamping transistors, it's a fairly standard topology used first in the mid 70s (Sansui AU517/717 come to mind).
Basically, it makes the VAS stage act as a complementary differential configuration, by extracting the oposite phase current of the LTP and by mirroring it around for the lower side of the VAS. It works even better if the diode is replaced by a suitable diode coupled transistor, i.e. a standard current mirror (of course, the transistors in it should be thermally coupled).