• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Tentlabs Neg Bias Vs Machmat AFB

Status
Not open for further replies.
Jump to Latest
J

jkeny

Member
Joined 2007
#1
#3
It's tough without schematics, but it looks like the Tent circuit is based on a sort of window detector (something like the autobias stuff that Bill Chater published) whereas the cheaper one seems to just be a servo with a long time constant.

CCS loading is fine for class A amps that never overload.
 
#4
SY,
Am looking at bias servo designs quite seriously at the moment for a parallel push pull 300B amp (Class AB1). The Tentlabs/Menno Vandervenne circuit is quite similar to the Morgan Jones circuit but rather than clip at 2 x the idle current to define a window around the bias point from 0 (tube cut off) to 2 x the idle current it uses a much tighter window around the required bias current point. It seems to be a good idea to limit bias drift when driving hard (into Class B).

I'm not aware off the Bill Chater stuff - any links (meanwhile I'll set Google to work).

I'm also looking at some of the John Broskie auto bias ideas which are very clever but less flexible (need to optimise his design for each application).

Cheers,
Ian
 
#8
It's official, they don't sell the cheap bias board anymore.

"I am very sorry, we do not sell them anymore. We got far too busy and could not deliver support anymore on them. We have a new version, microcontroller based, tunes the tubes on power (even better than current alone)m but we only use these in our modification, and for OEM, nor DIY anymore.

Mattijs de Vries, Machmat"
This was from a reply to an email I sent to the company.
 
#11
All valves have second-order distortion. As well as producing second-harmonic, this also gives a DC shift with signal. This DC shift is normal, and goes away as soon as the music stops. It doesn't change the required quiescent bias. Fixed bias maintains this. So DC up, quiescent current unchanged.

Cathode resistor bias, when bypassed, tends to oppose this and shifts the quiescent bias for a while until the capacitor discharges. So DC up (but less so than fixed bias), and quiescent current down. Some designs allow for this by deliberately setting the quiescent bias wrong, but knowing that signal will shift it.

CCS bias, or normal servo bias, is even worse. It strictly maintains the DC. So DC unchanged, quiescent down (even more than cathode resistor bias). I suppose you could allow for this by setting the bias even further wrong, but I would prefer to see correct bias for small signals to get minimum distortion at normal listening levels.

So the two 'correct' options are either fixed bias, or a special servo which only samples the current at signal zero crossings. I believe this is what the Tent circuit does.
 
#13
Bigun said:
Is it really necessary to set the bias so accurately, I thought tube circuits were rather forgiving (newbie question) ?
Click to expand...

I'm not sure how much accuracy is needed. The important qualities are keeping the bias stable through musical dynamics and correcting for shifting component characteristics (warmup, ageing, temperature) and maybe line voltage.

A very sophisticated controller might allow some offset at idle to offset a mismatch at higher signal levels. One might also argue that this is not the job of the bias controller and one should match tubes or balance the AC signal separately.

I've been thinking lately that auto-balance only might be a workable approach. It's not so important to hold the total idle current to a tight tolerance. Since it wouldn't account for line voltage, there would need to be a regulated B+ or circuit which is tolerant of B+ changes. The way B+ changes in my house, it's not clear I want the bias controller reacting to it in any case!

One point is what the bias controller does, how it works, might optimally be different if B+ is regulated vs. unregulated.
 
Last edited:
#15
Valve circuits are fairly tolerant, which is why most circuits can get away with cathode resistor bias. However, if the bias drifts too far off the correct value then you will get increased distortion - such as crossover in P-P. In most cases CCS or a simple servo is worse than a resistor, because it controls the wrong thing. It can be amusing to watch people spend lots of time and effort to make their circuit behave worse than it would with a simple resistor!

In small-signal circuits a CCS does no harm but it is not usually necessary as exact bias is less important than in output stages.
 
#17
Bigun said:
Ah - I hadn't thought of P-P as I'm still learning SE here (coming from SS world). I'm enjoying the fact that tubes are more forgiving, and glad to hear this is generally still the case.
Click to expand...

Hi

Yes, in SE it is less critical, though 2 equal channels for left and right are a must, in this aspect it still is critical, moreover because most tubeamps run with little to no gfeedback

In SE there's one exception; When using 2 tubes in parallel, one should make sure these are equally biased (and matched for gm).

best
 
#18
Hi,

Yes, in SE it is less critical, though 2 equal channels for left and right are a must, in this aspect it still is critical, moreover because most tubeamps run with little to no gfeedback
Click to expand...

Care to explain a little more indepth as I fail to see the relationship with bias offset in the PP amps?

BTW, I noticed Menno was using decoupled cathode resistors in his power amp.
That alone can be a source of distortion due to differences in ESR (the caps) and ordinary mismatch.
Subjectively, they tend to make the sound slow and the mask fine detail too.

Cheers, 😉
 
Status
Not open for further replies.
Top Bottom