Hiwatt Output Stages, NFB

[bea]

Ok, let me start with the Dynacords. I actually have four amps working

a) Dynacord Bassking 1 - like the Eminent 1 but only two inputs. First version with ECC808the input stages.

b) Dynacord Eminent 2 - "twice the Bassking"
Both with modified inputs - impedance increased. Coupling cap in input stage reduced to obtain to limit lower frequency limit to ~70 Hz 1st order.

c) Dynacord Eminent 2T - the widely spread hybrid version of the Eminent.

d) Dynacord G-2000 with DIY preamp. Again, the lower frequency response is limited to 70 Hz 6dB/Octave. This is meant to protect the speaker a bit from too low frequencies.

The cab in the picture at that time still ran the 2515 which i replaced for the 3015. It is tuned more or less like a TL606, i.e. to 56Hz in 90 l - measured by me. When i melted the 3015, the central port was closed, so the tuning was a bit lower. At that time I played the Gibson bass, which btw. has custom pickups, on the bridge pickup (but i do use the neck pickup quite a lot).

And yes, the red thing in the center of the speaker is a "beam blocker" - actually a hemisphere acting as a diffractor and set as close to the dust cap as possible without touching the speaker at maximum excursion.

Addition: the Oberton runs in 83 l at 56 Hz, and the Beta 12 runs in 55 l also at about 55 Hz


EDIT: i forgot to mention something - the preamp was strongly microphonic, especially when the gain pot was fully up. I could track it down to the valve (an EH ECC81).

BTW: the topic of a resonance in the preamp of the G-2000 is also on the table. The tube was using turned out to be strongly microphonic. Which i did not notice at home, at least to that extent - although i played the amp loud (at a time when that was permitted...).

[bea]

Ok, it looks as if i found something: i checked the output for a DC component. As long as the signal was small, the signals was clean. But as soon as the output level increases beyond a few Volts PP (about 4) a sudden galvanic "pulse" occured which quickly decayed within less than one second. That pulse remained invisible on the scope (AC coupled input). The amplitude was not too small - my analog volt meter showed up to 0.15 V (it must be larger - the mechanical inertia of the volt meter will dampen the displayed amplitude.)

Not good...

[J M Fahey]

Are we still talking the tubed Hiwatt ?

Then don't worry.

Transformers do not pass DC, period.

If your meter shows a negligible "DC value" in fact it is integrating the (pure AC) Audio signal, which if unsymmetrical (normal in a string musical instrument) will stay a little more on one side than the other or have a slightly higher average value one side than the other.

The display, which averages values along a certain time (or the display woukd flicker real time and be unreadable) or a needle which has a lot of mechanical inertia, takes that as a DC value which actually isn't there.

Your scope which does work in real time, sees nothing abnormal.

In any case, 0.15V does not tickle a mosquito.

[Mike Sulzer]

Ok, it looks as if i found something: i checked the output for a DC component. As long as the signal was small, the signals was clean. But as soon as the output level increases beyond a few Volts PP (about 4) a sudden galvanic "pulse" occured which quickly decayed within less than one second. That pulse remained invisible on the scope (AC coupled input). The amplitude was not too small - my analog volt meter showed up to 0.15 V (it must be larger - the mechanical inertia of the volt meter will dampen the displayed amplitude.)

Not good...

I am assuming that you were testing the amp with a slowly increasing sine wave and when you passed 4v out, a transient big enough to cause deflection on a mechanical dc meter occurred. That does not sound normal to me. Depending on how the scope was set, it could have missed it. Does your scope have a mode allowing a slow repetitive sweep where there is no erasure between sweeps? (I think this is sometimes called "envelope" mode, or something like that.) Sometimes that can catch transient breakdown or oscillation in circuits.

[J M Fahey]

Please get in perspective: we are trying to find what killed a 300W RMS speaker ; I wouldn't worry about what looks like 150 milli Volts DC and lasts one second .

If doubtful, just do the Math

[bea]

Juan, i did the math. But i am unsure and mentioned it.

To conclude: although the amp might show some minor signs of a former mistreatment, it looks as if it is in a usable state. What remains is what You told me about the capability of these eminence speakers to deal with even moderate actual loads. And that's pretty frustrating as a conclusion: a single 15" speaker is not capable to handle the power of even the G-2000 - or would high end speakers from the early 70s like the EV 15L or the JBL K-130 do?


@Mike: no my scope is an old russian 2 channel scope, fully analog and not in best condition. I bought it in a hurry when my old Hameg 203-5 died. I don't trust it for more than a few rough estimates - but anyway, it allowed me to finally finish the G-2000.

BTW: a picture showing the G-2000 on top of the Mywatt. To the left the Bassking on top of the Eminent.

[bea]

Anyway, i would like to understand what i see.

Maybe the following?

As i turned the pot quickly and so superimposed something like a step function to the signal and might see some charging/discharging effects of one of the energy reservoirs in the signal path, notably the coupling capacitors and the OT itself. In this power stage, there are three coupling capacitors within the NFB loop (input, cathodyn-PI, power tubes) which all have time constants to about 0.14 s (i.e. a corner frequency of about 7 Hz) which might fit to the time i saw this effect decaying.


Going back to the Hiwatt - i checked the corresponding time constants in the 200w Hiwatt plan and found much smaller time constants - 0.03 secs at the grids of the power tubes and roughly 50% longer at the PI.

Maybe it it advisable to reduce the coupling capacities in the Dynacord power stages as well ... i have to HiFi-Favorites with half the power waiting to be brought back to life, and my experience with the speakers might be an indication that it would be useful having an amp of that class with a hiwattisch preamp or something modeling its characteristics.

[Mike Sulzer]

I think turing the pot up fast should not change the energy storage unless you increased it enough to drive the amp non-linear. There should be no dc in the pots, and so only the ac should change. If you had class AB cathode bias, I suppose there could be an effect, but I think your power tubes used grid bias.

In general, an instability at the low end due to feedback effects seems unlikely. In any case you ought to hear it, for one of two reasons:

1. You are producing energy in the audible frequency range.

2. Even if the energy is outside the audible range, it would be taking a lot of power, and you would hear its effect on your signal.

There might be a sort of rule here: If your amplifier has an output transformer, and you destroyed the voice coil, then you just heard what did it.

I am not saying it holds all the time, but I bet it holds nearly all the time.

Maybe you could learn something without too high a cost by trying the light bulb limiter, and monitoring it. If the bulb is inside the box where it is dark and you bring out the wires from a suitable photo detector, I suppose that there is a smart phone app that would do the recording. Why not?

Anyway, i would like to understand what i see.

Maybe the following?

As i turned the pot quickly and so superimposed something like a step function to the signal and might see some charging/discharging effects of one of the energy reservoirs in the signal path, notably the coupling capacitors and the OT itself. In this power stage, there are three coupling capacitors within the NFB loop (input, cathodyn-PI, power tubes) which all have time constants to about 0.14 s (i.e. a corner frequency of about 7 Hz) which might fit to the time i saw this effect decaying.


Going back to the Hiwatt - i checked the corresponding time constants in the 200w Hiwatt plan and found much smaller time constants - 0.03 secs at the grids of the power tubes and roughly 50% longer at the PI.

Maybe it it advisable to reduce the coupling capacities in the Dynacord power stages as well ... i have to HiFi-Favorites with half the power waiting to be brought back to life, and my experience with the speakers might be an indication that it would be useful having an amp of that class with a hiwattisch preamp or something modeling its characteristics.

[bea]

Is there any possibility to split this thread?

We are actually talking about this output stage: http://www.peel.dk/Dynacord/pdf/G%20...7-01-1972).pdf

which Dynacord used in the Gigant, B1000, G1000, G2000, and a few others, the latest being the BA300. I removed all the semiconductors boards because it did not seem worthwhile to restore them (too many problems). I also omitted the limiter which was a feature unique to the G1000/G2000 and not present especially in the Gigant which is the most important and most wide spread of these amps.

The rapid change of the input signal causing the transient i observed is produced by a pot immediately in front of the input of the output stage.

That's how i came to my idea of the loading / unloading of the coupling resistors, which to my understanding always will occur during a sudden change of the signal amplitude.

My original idea was to use a Hiwatt input stage in the G2000, but i abandoned it in order to be able to use the original tandem pots, and just modeled the IMHO most important feature of the Hiwatt preamp - the position of the notch near 250 Hz. (And indeed, it sounds different, but actually really cool....)

[Mike Sulzer]

I see a push pull parallel EL34 amp driven by a 12AT7 split load (cathodyne) phase inverter, with the other section of the 12AT7 used as a stage of gain before the splitter. The pots I see just before the output stage are the bias adjustment pots. Am I missing something? (Yes, my vision is not as good as it used to be, and sometimes I miss things in busy schematics.)

If I understand the notation on the schematic, it takes about 70 volts p-to-p (25v RMS) to drive the output tubes, and I deduce that the phase splitter is able to do this, but does not have a lot extra drive.

It appears that the capacitor associated with cathode of the input stage of gain dos not affect the stability because both the open loop gain and the gain-around-the-loop would change together. The roll off of the transformer almost certainly is the dominant pole and the other two contribute their phase shift with the open loop gain below unity. So it appears that the low frequency NFB circuit is stable.

Is there any possibility to split this thread?

We are actually talking about this output stage: http://www.peel.dk/Dynacord/pdf/G%20...7-01-1972).pdf

which Dynacord used in the Gigant, B1000, G1000, G2000, and a few others, the latest being the BA300. I removed all the semiconductors boards because it did not seem worthwhile to restore them (too many problems). I also omitted the limiter which was a feature unique to the G1000/G2000 and not present especially in the Gigant which is the most important and most wide spread of these amps.

The rapid change of the input signal causing the transient i observed is produced by a pot immediately in front of the input of the output stage.

That's how i came to my idea of the loading / unloading of the coupling resistors, which to my understanding always will occur during a sudden change of the signal amplitude.

My original idea was to use a Hiwatt input stage in the G2000, but i abandoned it in order to be able to use the original tandem pots, and just modeled the IMHO most important feature of the Hiwatt preamp - the position of the notch near 250 Hz. (And indeed, it sounds different, but actually really cool....)

[J M Fahey]

FWIW I was also fooled in thinking we were still talking about a Hiwatt power stage, next time I won't pay that much attention at thread names

Wow!!! , a Gigant !!!!!
Gigant indeed, 200W RMS out of 4 x EL34 , better they are real Siemens/Valvo or Philips Eindhoven : 800V on the plates.

I used lots of Dynacord equipment when I was young, in neighbouring Paraguay, chock full of Naz** .... no, not really, that's the US/UK Gov't spread myth, simply a *strong* German immigration since the 1860's ... even their permanent President for 35 years was General Stroessner ... and every German stock teen traveled to Germany ... guess what brand amplifiers they brought back.

a single 15" speaker is not capable to handle the power of even the G-2000 - or would high end speakers from the early 70s like the EV 15L or the JBL K-130 do?

Search for the happy middle

Ultramodern Neos have a dissipation problem, not enough mass and surface to dissipate heat *long term* , old 70's speakers had obsolete adhesives and former materials; I bet the modern Overton , both with a massive steel and ceramic magnet *and* modern high temperature adhesives will not fail, at least not thermally.

Problem is weight, of course.

You *might* try to have a new cabinet made; a customer also bought an impossible to lift (no kidding, I can't) MB tube (6x6L6/5881) 200W , 2 x 10" EV type huge magnet speakers (probably Eminence made) and straight from the Factory they noticed that, so they provided built in rollers and an extensible strolling arm.

Couldn't find a good picture of somebody carrying one, so I post a few bad ones (there are different versions):

[Justin Thomas]

I'm sorry to derail this thread, but that just looks SOOOOOOooooooo much PANSIER than a proper head sitting on top of a proper cabinet, like Bea's. WTH, did Messa-Boogers buy out Roland and steal all their cabinet design features? It's even got the square metal grille on the front... Why don't they just cover it in black carpet and get the emasculation over with?

Look, I love art incorporated into amplifier cabinets, but that is just... "Hey! That Guy just stole my bass rig!" "You mean your luggage?" "No, my bass rig!" "Sir, I'm sure your luggage will arrive on the next flight..."

Justin

[J M Fahey]

You are not that far from the truth.
The real idea behind that actually very clumsy package (can't call it a "combo" ) was to sell it to *top* Pro session/backup players, think tha Bass player sharing a *huge* stagewith Madonna / Shakira / Luis Miguel / Ricky Martin / you-name-it.

Mind you, not a Rock band at all (nor rappers who have no use for this) but the Pro hires by some chintzy glittery singer.
Notice it's meant as a personal monitor, packaged in a flyable pack, which can be carried by owner through airports , across 5 star hotel lobbies, into expensive restaurants, TV studios , etc.

I just imagine Oprah interviewing a celebrity singer, "casually" asking him/her/h@ "this was not progranmed but .... "could" you sing just one song?" and him/her/h@ answering "ok, just by chance my musicians are here" or some similar populistic/PR cr@p

At least it gives the Bass player the (fake) sensation: "hey, here I am stuck inside a glittery tuxedo, playing this cr*p instead of some Heavy Metal ... but at least I am using a "TOOB" Mesa Boogie through a couple EVs"

Of course, later Reality kicks in: the Pro who bought it in Miami during a Tour playing with some Latin star sold it later because it was unusably heavy, and my customers, who plays Stadiums with 2 of my preamps driving an AB1500 and a QSC900 pushing 4 Ampeg 8x10" found that turning it on and off on stage was exactly the same, didn't change perceived sound level or quality in the least.

But the point is suggesting Bea she has a new cabinet made (or mod the present one) , with a good (heavy) speaker inside, but made carryable using these accessories which can be bought at

Penn Elcom - Flight Case and Speaker Cabinet Solutions

[bea]

No, You aren't, Mike. With the notable property Juan mentioned: 790 V on the plates.
Looking into detail, these power stages have few flaws - serious ones: the quality of the bias pots is poor, especially in the smaller version with only two power tubes (the Eminent).
In the Gigant power stage, there is insufficient head room to be able to bias all tubes (6CA7 built as tetrodes, KT77) . They simply doubled the bias circuitry when they went from 2 to 4 output tubes and did not account for the doubling of the current through R911 which also needs to be reduced to retain the bias range.

In all but the earliest versions of the Eminent power stage the NFB also supplies the ground of the cathode of the first tube stage!
(see http://www.peel.dk/Dynacord/pdf/Impe...)%20(1970).pdf)

I just started to restore two HiFi-Favorit 2 (2 EL34 @ 750 V, same power stage as Imperator) and plan to change this into the circuitry shown above.



Back to the 3015: it looks as if i was grossly in error. Last evening the speaker worked. For half an hour. I am not sure if it is fully ok - but there must be a flaky soldering point in my bass, and that of course seems to be the major problem. So it seems that "only" the 2515 has died. But anyway, i am waiting for my 2nd Oberton... and need to deal with the microphonics of the ECC81 in the G2000.

A remark on rolls and handles: useful unless You need to cross many stairs. I actually consider to supply some to my DIY cab.

[J M Fahey]

Last evening the speaker worked. For half an hour. I am not sure if it is fully ok - but there must be a flaky soldering point in my bass,

Not in yourBass but on the speaker itself.

Most suspect are ultraflexible tinsel/braided wires connecting fixed terminals on frame to *very much* moving voice coil terminals on the cone.
Way more so in a Bass speaker and worse if the cab is open in any way (including ports).