KT 88 project request

[Gingertube]

Just for info/guidance:
A few years back I did a 4 x KT88 into Hammond 1650T Ultralinear HiFi Power Amp.
I used 360-0-360 HV power tranny which gave me a B+ of a few volts over 500V,
Amp delivered 120W continuous sine wave with 135W in heavy overdrive.
I used a Curcio Audio Engineering cascode diffamp front end.
This one, which is intended for a Dynaco MKIII Upgrade:
It has become widely known that the output transformers in the Dynaco ST
THE DUD120 name was a bit of fun DUD for "Down Under Designs"
Cheers,
Ian

[Fred Von Beerdrinkelburg]

Hi!
I've been looking for the same thing as the catalin and I'm extremely tempted to start ordering parts to your schematic Gingertube!
I'm just wondering about the pt.. having a hard time finding one with fitting secondaries, any suggestions?

How's your project comming together catalin?

[catalin gramada]

Hi.Little changes from initial plan...KT90 and EL84 triode connected as driver.Just arrived at home so was no time to wire something yet.Just P'S was tested. 590V without load.hope I can benefit 560V under load.
My chassis
Gingertube schematic looks more smart. I choosed to keep it simple instead and use brute force cause I have a lot of resources. power tranny 500w and 500uF main filtering
That.s a really big engine. I can push two amps like that

[catalin gramada]

Hallo.
Almost finished my kt90 project. El84 triode connected drivers 280v on plate from 560v supply on 10k plate load. Have not recorded any issues at all but heating inconvenience for 28mA per side.


I have one question for experts.please.
I choosed to use a baxandall tone control after first voltage amplifier drived directly from 12ax7 plate. I did some experiments before with cathode follower and I don.t like the results. Plate driving sound better to my ears but it is too much dependant by volume pot.I found it more sensitive than a classic Fender tone stack from this point.The further stage it is a 12at7.How can I fit better a baxandall circuit with a volume pot.please? Thanks.Catalin

[KevinOConnor]

Hi Guys

The ideal Hammond OT match for the 278CX is the 1650TA. With any four large-bottle power tubes you will get 160W.

If you begin with the basic plexi-style circuit and the standard MI power amp, there is no problem and you have a good basic sound. The standard PA uses fixed screen voltages, but you can easily put in a switch or hard-wire it as UL. Remember to leave the screen-stops intact in ALL operating modes. There is no need for further amplification and you end up with something that is similar to the Sunn, or like a "baby" Major.

Regarding the heater ratings: Hammond suggests that adding the usual three preamp tubes is no concern on the heater winding. They used to consider anything up to 25% over-load to be safe, and the 900mA is only 15%.

You should also try the UL stage open-loop. It has a natural compression effect that is great with guitar and bass. All the guitar amps I've built have been like this, except for the Studio amp which has a feedback loop that can be dialled in/out and several operating modes for the output stage. UL has generally been misued by the big name amp builders, so players disregard it.

Definitely stay away from cathode-biasing if you want serious power. It is fine to have switchable bias where the cathode setting is for much less power and just a different tone. In which case, Rk should be unbypassed to highlight the tonal difference.

Have fun

[catalin gramada]

Hey Kevin. I know what you mean.My project use 1650t and pair of kt90 in ul mode at 580v in the best days.Drived from pair of el84 triode mode and have a 12at7 split load inverter.Independent supply for power stage and preamp and bias.No compensation networks included.It.s a beast.Punch and definition and sound stage.whatever you want from an amp. But I.m not familiar with signal filtering networks.how to integrate it into an amp . I.m tried to included a tone corector in front. A baxandall after first voltage amplifier and feel a lacks of details like lose some transparency. I expect just to shape the tone not to flat the sound image killing transients. I thought is a impedance mismatching and tried a cf.but I did not liked the result. Which is correct way to integrate a tone control between stages.like general conditions.principles.please? Thanks

[KevinOConnor]

Hi Guys

A CF-driven EQ is a mistake for MI, especially if the EQ is the usual tone "stack" - the EQ will be ineffective and lifeless. Bypassing the CF, the stack can provide a dynamic load to the preceding gain stage and the EQ comes alive.

Although I like the Baxandall EQ and have used it in several products, simpler tone control circuits often work much better for guitar and bass. Every passive EQ introduces a signal loss even when set to full bass and treble gain. A Bax at least has a true flat response when the pots are set mid way. Anyway, these passive EQs are really just cut and more cut and there needs to be "make-up" gain afterwards.

With guitar, there are lots of mid and high frequencies that we can hear, and the circuit adds harmonics to emphasise things further, making it very easy to hear what the EQ is doing. For bass, things are more difficult except for the higher notes or if some distortion is allowed from the preamp, and this may be just a slight fattening or compression effect but it makes the EQ easier to hear.

Many tube bass preamps only have one tubestage or at most two and usually one is bypassable. The high output of a bass guitar means that very little gain is needed to drive the power amp to full output. However, each tube stage adds a viel of character to the sound and can make the sound more "effortless" - sometimes more of something you feel than hear.

The circuit you described above has a gain stage, Bax EQ, presumably a second gain stage, the concertina splitter, the EL-84 driver stage, then the output tubes. There should be enough gain there to accommodate the EQ loss, but typically, there would be a dedicated make-up gain stage after the EQ, then the power amp front-end gain stage. It is usual with the concertina to use the other triode in the bottle as the PA front-end gain. Overall, there would be a 2-stage preamp, then your effectively 3-stage PA.

The engineering ideal is to drive EQs from a low-impedance, but this is usually because the EQ may dip to a low-z itself and load the signal source. But the "ideal" is not always the same as what we need or what sounds best, so assuming you are using typical high-Rs and low-Cs in the EQ, there is no need for low-z drive. Use that CF as another gain stage and set its gain to be low, and/or insert another level control or interstage attenuator driving it. This will give you more degrees of freedom for voicing the amp.

Have fun

[catalin gramada]

Thanks for you kind reply. Can be possible to have more benefit from the same corrector moving it between second gain stage (means first half of 12at7) and phase splitter.(second half 12at7) Please? Thanks.
Catalin

[KevinOConnor]

Hi Guys

Just so everyone is clear, the split-load stage (concertina) IS the splitter and it has no gain of its own. The stage driving this one is a separate entity from the splitter even if it is direct-coupled to the concertina.

There can be advantages having the EQ later in the circuit, and it should be easier to hear what it is doing later rather than early. This is especially so in an amp that is set up primarily as a clean machine. As I said before, each stage adds harmonics. The first stage adds a bunch and passes everything along to the second stage. The second stage adds harmonics to the amplified signal plus to the harmonics added by the first stage. So, the output of the second stage is much richer with harmonics than the output of the first stage, and this will give the EQ more to work with. This is exactly why in a guitar amp with a distortion channel, it is most common to have late-EQ, as early-EQ becomes less effective if most of the distortion is added after it (as in a Dumble amp).

Have fun

[catalin gramada]

...so that means more current through the EQ driver and not compulsory low ra EQ driver?

[KevinOConnor]

Hi Guys

If by "driver EQ" you are thinking of placing the EQ between the EL-84 drive stage and the output stage, then don't do it! That is a circuit point where significant signal swings are present and required to achieve the full output of the amp. Were an EQ inserted here the drive stage would be called upon for even greater output than it already produces and possibly more than it can.

With typical values in the EQ, the current through it is of no concern and does not require a CF to drive it. Look at every Fender amp other than certain Bassman models - plate-driven EQ and EQ that does something.

If you have a scope, it is easy to see the signals needed at each circuit point to achieve full output. This is the kind of test you do with a bench load not a speaker. You can do it without a scope but it is a bit more difficult. In any case,, you can see where there might be a "signal bottle neck" suggesting that such a point is NOT where an EQ should be placed. In a typical circuit, the EQ is placed after the first or second stage. It is unclear exactly how many gain stages your amp has.

Have fun

[catalin gramada]

It is as you guess first voltage amplifier. second voltage amplifier.phase splitter.power drivers.and power tubes.Thinking to play more to the voltage of second gain stage and bias to the left side to rise it to max power dissipation value and to use this stage to drive the EQ. Should be more convenient than to use a large bypass cap over rk to low it's impedance. Bypassed I don't like how respond in bass anyhow.more expansive and somehow uncontroled over all frequency range (or maybe more undampened it is a better term. too deep and vibe from an aesthetical point)

[KevinOConnor]

Hi Guys

In that case, the usual position for EQ is where you have it, after the first stage. Depending on how much drive is heeded by the EL-84 stage, you could try placing some EQ after the second stage feeding the splitter.

Note that there are simple EQ circuits that can roll off bass or roll off highs, that use just one pot and one cap. The latter looks like a guitar tone control and can be placed anywhere. The former is like the 'shape' control in my Soma-84 amp in TUT5, where a second coupling cap is placed in series with an existing one and a pot is paralleled with this new cap. The new cap is a much lower value than the existing plate cap, so when it is in play, the effect is to roll off bass. The pot shorts out the small cap at one end of its sweep, providing full bass.

Similarly, if there are any stages that do not have cathode bypass caps, one can be added in series with a pt to dial it in. as a treble boost or straight gain boost.

The topology as described adds up to three triodes plus the EL-84s and outputs, leaving one triode unused? That could be added ahead of the splitter giving more places for the EQ and more points to provide voicing for the amp. That is what I referred to earlier.

Have fun

[frus]

A CF-driven EQ is a mistake for MI, especially if the EQ is the usual tone "stack" - the EQ will be ineffective and lifeless. Bypassing the CF, the stack can provide a dynamic load to the preceding gain stage and the EQ comes alive.

You mean like the tone stack has more range if driven directly from gain stage?

[Gingertube]

YES,
Kevin can respond for himself BUT this is my take on it.
The load on the triode stage is the anode load resistor in parallel with the impedance (to signal ground) of the tone stack it drives.
The impedance of the tone stack varies with frequency according to the settings of the TREBLE, BASS and MID pots.
That means that the load of the driving triode stage varies with frequency according to the pot settings.
That means the gain of the driving stage varies with frequency, or if you like, the signal at the anode varies with frequency BEFORE it hits the tone stack itself. It is then further modified by the tone stack.
It is much more interactive than a CF driven stack.
Cheers,
Ian